The Strange Death of Jeffrey Epstein, Revisited

Six Years Later, The Investigation That Refuses To Stay Closed, A Forensic Review Incorporating The 2025 Document Releases, The Edited Surveillance Video, And What Still Cannot Be Reconciled

I have been asked more times about this case than about any other I have touched in my career, and for years I resisted writing a second piece, because the first one, published on rauscher.xyz in early 2025, had already stated what I thought needed stating. Then the documents started coming out. The Department of Justice memo in July 2025. The surveillance video that was described as full raw footage and which forensic analysts then demonstrated was anything but. The Wired investigation into the file metadata. The CBS News investigation into what the video actually shows, and what it does not show. The House Oversight Committee's release of the Epstein estate files in September 2025. The Epstein Files Transparency Act, signed into law on November 19, 2025, and the Department of Justice's partial and heavily redacted December release that triggered a bipartisan senate letter accusing the department of violating its own compliance deadline. The case changed substantially in 2025 in ways the original article could not have incorporated, and several of my earlier observations, including one about a cosmetic detail that I should have treated with more caution, deserve correction in light of material I had not fully considered.

This is that correction, and that update, and that deeper forensic look, written at the end of April 2026 with the benefit of everything the past fifteen months have produced.

What Changed in July 2025

On July 7, 2025, Axios published a two-page unsigned memorandum jointly issued by the United States Department of Justice and the Federal Bureau of Investigation, concluding after what the document described as an exhaustive systematic review that Jeffrey Epstein had died by suicide at the Metropolitan Correctional Center in Manhattan on the night of August 10, 2019, that no so-called client list existed in any searchable form within the government's holdings, that no credible evidence of blackmail against prominent individuals had been identified, and that no predicate for investigating uncharged third parties had emerged from the case files (Axios, 2025, Exclusive: DOJ, FBI conclude Jeffrey Epstein had no client list, committed suicide). The memo stated that investigators had conducted digital searches of databases, hard drives and network drives, along with physical searches of cabinets, desks and closets, to identify any and all relevant materials within federal custody, and that the conclusion aligned with the determinations previously reached by the Office of the Chief Medical Examiner of New York City, the United States Attorney's Office for the Southern District of New York, and the Department of Justice Office of the Inspector General (DOJ/FBI Joint Memorandum, July 7, 2025).

Attached to the memorandum, and available through the department's public website, were two versions of what the agencies described as the full raw surveillance recording from the Special Housing Unit of the facility where Epstein had been held. One was unaltered, the other enhanced for contrast and clarity. Taken together, the agencies said, this footage captured the relevant common area from approximately 10:40 on the evening of August 9 through approximately 6:30 the following morning, and would have recorded anyone entering or attempting to enter the tier where Epstein's cell was located.

This is the part of the official record that subsequent forensic analysis has effectively unraveled.

The Video That Was Called Raw And Was Not

Within days of publication, the technology publication Wired conducted a forensic analysis of the video file the Department of Justice had released, examining the metadata embedded in the file itself, and published findings that contradicted the government's description of the material in several substantial respects (Wired, July 15, 2025, Metadata analysis of Epstein surveillance file). The embedded file history showed that the recording had been opened and edited in Adobe Premiere Pro, a professional video editing application, by a Microsoft Windows user account identified in the metadata as MJCOLE approximately twenty-three minutes before the file was saved and exported. The edit history indicated that the final file was assembled from at least two separate source clips, saved four times across the editing session, and exported in a process that is structurally incompatible with the term full raw in any professional or forensic sense of that phrase. The creation timestamp on the file was May 23, 2025, approximately six weeks before the Department of Justice released the material as the unedited video record of Epstein's final hours.

CBS News conducted its own parallel forensic review using several independent video specialists, including the digital forensics expert Hany Farid, whose analysis concurred with the Wired findings and extended them. Farid, whose commentary on the CBS review was unusually blunt for the normally restrained world of forensic expert witnessing, told the network that if a lawyer had presented him with the released file and asked whether it was suitable for court use, his answer would have been no, and that the only appropriate remedy was to return to the source system and perform a proper export without editing intervention (CBS News, July 28, 2025, Jeffrey Epstein jail video reveals new discrepancies). The CBS team also confirmed, through a high-level government source, that the FBI, the Federal Bureau of Prisons, and the Department of Justice Office of the Inspector General each maintain copies of the original surveillance footage in which no one-minute gap exists at the boundary between August 9 and August 10, contradicting the explanation offered at a July 8 Cabinet meeting by Attorney General Pam Bondi, who had attributed the visible missing minute to a nightly reset of the outdated prison security system.

Attorney General Bondi's explanation was, strictly speaking, testable, and the test was conducted by forensic analysts who had no institutional stake in the outcome. The test failed. The nightly reset hypothesis implied that every twenty-four-hour cycle in the prison's surveillance system would show a corresponding minute-long gap. Video specialists reviewing the archive found that no such regular gap appeared elsewhere in the available material. The absence of a minute at precisely the boundary between the two calendar days during which Epstein died was not a structural feature of the system. It was an artifact of the edit, performed in Adobe Premiere Pro on May 23, 2025, by a user whose account name has been documented but never officially explained.

On September 2, 2025, the House Oversight Committee released a congressional video drop containing approximately thirty-four thousand files from the Epstein investigation archive, which included two clips that together filled the previously missing minute between 11:58:59 on the evening of August 9 and midnight on August 10 (Fox News, September 2, 2025, Missing minute from Epstein prison cell). When Fox News Digital combined the two clips, the resulting stitched footage demonstrated that no gap had existed in the original prison recording. One of those source clips had been trimmed by two minutes and fifty-three seconds from its original length, and the aspect ratio of the footage after the midnight boundary differed slightly from the footage before it, both indicators that a substantive edit had been performed on source material that otherwise would have run continuously.

The sequence of events describes itself. The government released a file it called raw. Independent analysts demonstrated the file was not raw. A congressional committee then released the source material that contradicted the government's release. The attorney general's public explanation for the anomaly was contradicted by both the forensic metadata and by the behavior of the underlying surveillance system. And none of this happened in secret. It happened across the July-through-November window of 2025, documented in real time by two of the largest news organizations in the United States and by the congressional oversight committee with jurisdiction over the Bureau of Prisons.

Whatever one concludes about Epstein's manner of death, the integrity of the evidence chain produced by the executive branch has been substantively compromised in a way that, in any courtroom I have testified in during my twenty-five years of forensic practice, would have disqualified the material from being considered reliable primary evidence.

The Cell Door That Was Never On Camera

There is a detail in the CBS News investigation that received less public attention than the missing-minute controversy but which, from a forensic standpoint, is considerably more consequential, because it speaks to what the surveillance record was actually capable of showing in the first place, regardless of what edits were or were not performed on the file later.

The released footage, cross-referenced with the floor plan of the Metropolitan Correctional Center and with diagrams of the Special Housing Unit, does not show Epstein's cell door, does not show the main entrance to the Special Housing Unit, and does not show the staircase leading up to his cell in full. What the footage shows is a stairwell visible only in a narrow sliver, a desk area, and a portion of the common area of the Special Housing Unit, with movement in and out of frame at the edges of what the camera could capture (CBS News, September 22, 2025, Jeffrey Epstein jail video reveals new discrepancies).

Julie K. Brown, the Miami Herald reporter whose 2017 and 2018 investigative work broke the Epstein story internationally and who probably knows the facility layout better than any other journalist in the United States, stated on CNN in late July 2025 that there were multiple other entrances to Epstein's wing that the released surveillance video does not cover, and that it was plausible that someone from another part of the area could have accessed his cell without appearing on any of the footage the government has made public. In Brown's words, the doors visible in the released material are not the doors that matter, because none of them are the door of his actual cell (Daily Beast, July 29, 2025).

The specific claim the Department of Justice and the FBI made in their July 7 memorandum, that anyone entering or attempting to enter the tier where Epstein's cell was located from the Special Housing Unit common area would have been captured by this footage, contains an important qualifier that became audible only after the floor plans were examined against the video. The phrase from the SHU common area is load-bearing. It does not claim the footage captured all possible entries to the cell. It claims the footage captured entries that proceeded through one specific route. Other routes existed. Those routes were not recorded, or if they were recorded, the recordings have not been publicly released.

CBS News's analysis identified several additional inconsistencies that had never been publicly addressed. Multiple individuals are visible in the released footage entering and exiting areas that, according to the statements of correctional officers Tova Noel and Michael Thomas given during the federal investigation, should have been inaccessible without either a remotely operated door or a physical key that Noel testified was in her possession at the relevant times. The video shows these entries occurring at times when neither officer was near the door in question or, in some cases, not present in the unit at all. Either their statements are incorrect, or the access controls at the facility functioned differently than the officers described, or both. No subsequent investigation has clarified which of these possibilities is true.

When Attorney General William Barr stated in August 2019, after conducting what he described as a personal review of the surveillance video, that no one had entered the area and that the medical examiner's suicide determination was supported, he was stating a conclusion about footage that, as we now know six years later, did not show the area that would have needed to be shown in order to support that conclusion. Whether Barr's statement was made in good faith based on inadequate technical information, or whether it was made with knowledge of the coverage gap, is a question I cannot answer. What is clear is that the claim itself, as a forensic statement about what the surveillance system recorded, cannot be sustained on the basis of the material the public has actually been shown.

The Ligature That Did Not Match the Wound

This is the detail I want to address with the most care, because it is the detail that has moved the most decisively from dispute toward technical clarity in the years since Epstein's death, and because it is the detail most directly within my own professional training.

When a body is recovered from a suspected hanging, the forensic pathologist examines the neck injury with reference to what is called the ligature furrow, the depression or discoloration on the skin that corresponds to the object that caused the constriction. The characteristics of that furrow, its width, its depth, its angle, its continuity around the neck, the distribution of associated subcutaneous hemorrhage, provide information about the object that produced it, and about the manner in which that object was applied. Hanging by a soft, folded bedsheet produces a ligature furrow with specific properties. The furrow is broad, because the folded sheet spreads the constriction over a substantial width of contact area. The furrow tends to ascend toward the point of suspension, following the vector of the gravitational pull. The edges of the furrow are soft rather than sharply demarcated, because the sheet itself has no hard edge.

Ligature strangulation, by contrast, tends to produce a horizontal or nearly horizontal furrow, because the application of pressure is not gravitationally directed but is actively applied around the circumference of the neck. The furrow tends to be narrower if the ligature material is thin, and the edges tend to be more sharply defined. Associated findings can include petechial hemorrhages in the conjunctivae of the eyes, in the oral mucosa, and in the skin of the face, because the cessation of venous return from the head while arterial supply continues, even briefly, produces the burst capillary pattern that forensic pathologists have documented for well over a century (Ligature mark on neck: how informative, Journal of Indian Academy of Forensic Medicine, 2005, 27(1), 10-15).

The photographs of Epstein's ligature furrow, released by his brother Mark Epstein during the post-autopsy period and subsequently reviewed in detail by the independent forensic pathologist Michael Baden, show a wound that is narrow, concentrated, roughly horizontal, and not obviously consistent with the broad, soft furrow that a folded orange bedsheet would be expected to produce. The photographic evidence released by CBS News in January 2020 showed the alleged ligature, a folded and twisted sheet, in condition that was, according to Dr. Baden's assessment, too wide and too smooth to have produced the wound seen on Epstein's neck. In Baden's direct quotation to CBS, the ligature found in the cell doesn't match the ligature mark on the neck, and specifically it is too wide, too smooth, and, critically, bore no transfer of blood or skin from the neck onto the fabric, despite the wound showing that the ligature had been in tight contact with the neck for an extended period before the body was discovered (CBS News, January 2020, Jeffrey Epstein's autopsy: a closer look, 60 Minutes).

The absence of blood or tissue transfer onto the alleged ligature is a specific forensic finding that deserves to be stated more plainly than it has been stated in the public discussion. A ligature that has been in tight contact with a bleeding injury on the neck of a victim for hours, as the body's lividity pattern indicated the ligature had been, should show physical evidence of that contact. The fibers of the fabric should hold blood residue. The texture of the fabric should show skin cells and microscopic tissue transfer. The testing for such transfer is routine. The existence or non-existence of such evidence on the sheet recovered from Epstein's cell has never been publicly disclosed by the medical examiner's office, and repeated requests for the forensic inventory have not produced an answer. The Connecticut Examiner published an opinion in February 2026 making the same point from a different angle, noting that when autopsy findings do not match the scene, forensic pathologists trust the body rather than the scene, and that the injury pattern in Epstein's case was uncommon in routine suicidal hanging to a degree that could not be reconciled with the physical evidence recovered from the cell itself (Connecticut Examiner, February 17, 2026, Epstein's autopsy deserves a second look).

The hyoid bone fracture, which received most of the early attention, remains relevant but is not, on its own, definitive. Hyoid fractures do occur in suicidal hangings, particularly in older individuals, though the rate has been estimated across multiple published studies as approximately six percent of hanging suicides, compared with approximately seventy percent of homicidal ligature strangulations (Khokhlov, 2001, Calculation of tolerance limits when analyzing the occurrence of fractures of hyoid bone and laryngeal cartilages in hanging, Forensic Science International, 121(1-2), 32-34). What has received less attention is that Epstein's autopsy did not show a single hyoid fracture but rather multiple neck bone fractures, including the hyoid and at least two additional cervical structures. The probability of multiple cervical fractures occurring in a suicidal hanging from a low bunk, which is the scene the authorities described, is substantially lower than the probability of a single fracture, and the multiple-fracture pattern aligns more closely with the injury profile of active manual strangulation or ligature strangulation than with the injury profile of passive gravitational hanging.

The petechial hemorrhaging in the sclerae and conjunctivae of Epstein's eyes, documented by Baden during the autopsy he observed, is another finding that points away from a textbook hanging suicide. Petechiae arise when venous return from the head is obstructed while arterial supply continues, even briefly, and the resulting back-pressure bursts small capillaries. This pattern is characteristic of strangulation, where the ligature or manual pressure occludes veins while allowing partial arterial flow. In hanging suicides with a high drop, the airway is typically closed rapidly and completely enough that the characteristic petechial pattern does not develop. In a ligature situation in which arterial and venous occlusion are unequal, the petechiae develop clearly.

The physical evidence, taken together and viewed six years later with the benefit of all forensic commentary that has since been published, describes a death that is more consistent with ligature strangulation than with suicide by hanging from a soft folded bedsheet. The alternative implement, whatever it was, has never been identified publicly. The sheet recovered from Epstein's cell shows no evidence of having been in the contact that the wound pattern on the body indicates occurred. Those two findings are not reconcilable with the official narrative as it has been stated.

The Body On The Gurney, And A Correction I Owe

In my original article, I discussed the widely circulated post-mortem photograph of Epstein being transported at the Manhattan hospital, and I stated, based on an ear comparison that I performed using the high-resolution reference material available to me, that the person on the stretcher was not Jeffrey Epstein. I further suggested that the most parsimonious explanation for the apparent discrepancy, if the body was indeed not Epstein, was that the death had been staged. I want to revisit that analysis in light of material I did not fully consider at the time.

Ear identification, which is my area of forensic specialization, is based on the individual variability of the auricular structure, including the morphology of the helix, the antihelix, the tragus, the antitragus, the concha, and the lobule, along with the specific relationships between these structures. The claim that ears are as unique as fingerprints is well established in the forensic anthropology literature, and I stand by the ear analysis itself as a technical exercise. Where I need to be more careful is in the interpretation of what observed differences might mean.

Two factors complicate the body-double interpretation that I offered in February 2025. The first is a point about nasal tip morphology that I want to address directly, because I referenced it in the original article as part of the evidence pointing toward a different individual on the stretcher. The lowered and somewhat flattened nasal tip visible in the post-mortem photograph, compared to the more elevated and defined nasal tip visible in earlier photographs of Epstein while alive, is consistent with an identifiable pattern that I should have weighed more heavily. Hyaluronic acid fillers, injected into the dorsum and tip of the nose for the purpose of producing a refined, elevated, and more youthful nasal appearance, are widely used in the cosmetic procedures of affluent and image-conscious populations, and were especially common in the circles in which Epstein moved during the decade preceding his incarceration. The conventional assumption that these fillers fully dissipate within three to twelve months has been substantially revised in the recent peer-reviewed literature. Master and colleagues, publishing in Plastic and Reconstructive Surgery Global Open in 2024, conducted an MRI-based study of thirty-three patients who had received hyaluronic acid fillers in the midface region, and found filler presence in all thirty-three patients at varying intervals post-injection, with some patients showing filler presence up to fifteen years after the most recent documented injection (Master, 2024, Hyaluronic acid filler longevity in the mid-face: a review of 33 magnetic resonance imaging studies, PRS Global Open, 12(7)).

What this means in practical terms is that filler material introduced for cosmetic nose refinement can persist for years while simultaneously undergoing gradual volumetric reduction, partial migration, and structural change as the cross-linked polymer breaks down and is reabsorbed by the body. The visible result over time can include a loss of tip projection, a flattening of the previously lifted contour, and a subtle shift in the profile that is particularly pronounced when the surrounding skin and underlying muscular support have also aged. In a man of Epstein's age, sixty-six at the time of death, with a several-year history of incarceration-related stress, weight fluctuation, and reduced access to the cosmetic maintenance routines he had reportedly maintained in earlier years, a pronounced change in nasal tip morphology is substantially more likely to reflect filler degradation than identity substitution. I should have given this explanation the weight it deserved before offering a more dramatic interpretation, and the correction is owed.

The second factor is that the New York City Office of the Chief Medical Examiner identified the body through fingerprint comparison against Epstein's fingerprint records, a procedure that operates at a level of individual discriminability that is not subject to the same degradation variables as surface photographic comparison. The identification of the body as Epstein is, on the basis of the fingerprint record, essentially unchallengeable. The body on the gurney was Epstein. The ear differences that I observed, to whatever extent they were real rather than artifacts of lighting and angle in the photographs, do not override a fingerprint identification, and I should have said so more clearly in the earlier article.

This does not rehabilitate the official narrative about Epstein's manner of death. It removes one of the more exotic theories from the table, leaving the more serious and better-supported questions about the ligature, the injury pattern, the video record, and the institutional handling of the case intact and, in several respects, strengthened by the 2025 disclosures. The question is not whether the body was Epstein's. The body was Epstein's. The question is how Epstein came to be dead in the condition in which the body was recovered, and that question remains unresolved.

The Guards, The Cellmate, And The Chain of Procedural Failures

The institutional context around Epstein's death has been examined in greater detail by the Department of Justice Office of the Inspector General than any other aspect of the case, and the 2023 OIG report, followed by the 2025 DOJ/FBI memorandum, represents the official account of what went wrong at the facility. The baseline facts are not disputed. The two officers on duty in the Special Housing Unit, Tova Noel and Michael Thomas, failed to conduct the required thirty-minute inmate checks throughout the night of August 9 into the early morning of August 10. Evidence developed during the federal investigation indicated that both officers were either asleep or browsing the internet, with one officer shopping online for furniture and motorcycles during the period that Epstein was, according to the official timeline, dying in his cell. Both officers subsequently falsified the log entries to indicate that the required checks had been conducted, and both were charged with federal offenses related to the falsification, entering into a deferred prosecution agreement that required admission of the falsification but avoided custodial sentences in exchange for probation and community service (AP News, November 2019; Reuters, December 2021).

Epstein had been removed from formal suicide watch on July 29, 2019, six days after the first incident in which he was found on the floor of his cell with marks around his neck, a period of observation that former Bureau of Prisons officials described as unusually brief for a high-profile inmate who had either attempted self-harm or been attacked by a cellmate, depending on which account of the July 23 incident is accepted. The cellmate who had been assigned to share Epstein's cell after his removal from suicide watch, a former police officer awaiting trial on homicide charges, was transferred out of the cell the day before Epstein's death, and no replacement cellmate was assigned, leaving Epstein alone in the cell during the critical overnight period, in violation of the facility's protocol for inmates recently removed from suicide watch (New York Times, August 2019; Politico, June 2023).

The 2023 OIG report described this accumulated pattern as a combination of negligence, misconduct, and job performance failures that created the environment in which Epstein's death was allowed to occur. The report did not conclude that the pattern was evidence of deliberate facilitation of a homicide. It concluded that the pattern was evidence of institutional dysfunction severe enough to permit a self-inflicted death that would otherwise have been prevented by standard operating procedures.

There is a reading of this conclusion that I find persuasive in part, and a reading that I find inadequate in part. The pattern of failures was indeed severe. The pattern was also, in its combination, extraordinarily convenient for any interested third party who might have had reason to ensure that Epstein did not survive to testify. The removal of the cellmate. The failure of the cameras directly outside the cell. The inattention of the guards during a specific three-hour window. The absence of any inmate companion who might have witnessed what happened. Each of these individual failures has a defensible institutional explanation. The probability of all of them occurring simultaneously, during the specific hours when Epstein died, against a background of federal charges that threatened to expose the activities of a substantial number of extraordinarily wealthy and powerful associates, is a statistical question that institutional investigators have not engaged with directly. A forensic investigator asked to evaluate the probability of this specific combination of failures occurring in the absence of external coordination would, at minimum, want to see the alternative hypothesis tested rigorously. That test has not been conducted publicly, and based on the 2025 DOJ/FBI memorandum, it is not going to be.

The Epstein Files Transparency Act And What It Has, And Has Not, Produced

On November 18, 2025, the United States House of Representatives passed the Epstein Files Transparency Act by a vote of 427 to 1, with Republican representative Clay Higgins of Louisiana casting the only dissenting vote. The Senate passed the bill by unanimous consent the following day, and President Donald Trump signed the bill into law on November 19, 2025, creating Public Law 119-38 (Congress.gov, 2025, H.R. 4405; White House Statement, November 20, 2025). The statute required the Attorney General to make publicly available in a searchable and downloadable format all unclassified records, documents, communications and investigative materials in the possession of the Department of Justice that related to the investigation and prosecution of Epstein, within thirty days of enactment, and required within fifteen days of that release a report to the Judiciary Committees of both chambers listing all categories of records released and withheld, a summary of redactions made with their legal basis, and a list, without permitted redaction, of all government officials and politically exposed persons named or referenced in the materials.

The deadline for the complete release was December 19, 2025. The Department of Justice did release documents on that date, but the release was partial, extensively redacted, and in the assessment of a bipartisan group of twelve senators including Richard Blumenthal, Lisa Murkowski and Jeff Merkley, materially non-compliant with the statute the Congress had just passed. The senators' December 24 letter to the acting Inspector General described the department's release as containing documents that were already publicly available, as withholding files within the scope of the statute, and as applying redactions to an extent that raised serious questions about whether the department was properly applying the limited exceptions permitted under the Act (CNBC, December 24, 2025, Epstein files: DOJ says one million new documents found).

The department's compliance with the Transparency Act is ongoing, and more material has been released since the initial December drop, but the pattern of partial disclosure, followed by congressional and journalistic documentation of omissions, followed by further tranches released only after public pressure, has repeated itself through the early months of 2026. The House Oversight Committee's release in September 2025 of approximately thirty-four thousand files obtained through subpoena from the Epstein estate, which included the now-widely-discussed birthday book compiled for Epstein's fiftieth birthday in 2003, was significant not only for the prominent names and the controversial contents but for the fact that this material entered the public record through congressional rather than executive branch action. The executive branch, across two administrations and under attorneys general of both parties, has consistently been the party constraining the flow of information, and the congressional and judicial branches have consistently been the parties forcing material into the daylight.

Whatever one concludes about the specific circumstances of Epstein's death, the pattern of institutional behavior around the case is documentable, non-partisan, and consistent: the agencies that possess the information have, for six years, resisted disclosing it, and the mechanisms that have succeeded in extracting it have been statutory, judicial, and journalistic rather than voluntary. This is not a conspiracy theory. This is a behavioral pattern in the public record.

What Remains Undecided And What Does Not

There are several matters in this case that I consider, after revisiting the full record, to be effectively settled, and several that I consider to remain genuinely open.

The matters I consider settled include the identity of the body, which was confirmed by fingerprint comparison and which I no longer believe should be questioned on the basis of photographic analysis of secondary features such as nasal profile or ear morphology, particularly given the alternative explanation that filler degradation provides. The matter of the guards' behavior during the night of August 9 to August 10 is settled: they failed to perform their duties, they falsified the logs, and they acknowledged both actions in their deferred prosecution agreements. The matter of the institutional failures at the facility is settled: multiple simultaneous breakdowns of standard procedure occurred, creating a window during which Epstein was unobserved.

The matters I consider genuinely open include, most centrally, the manner of death itself. The physical findings from the autopsy, including the multiple neck fractures, the petechial hemorrhaging, and most pointedly the mismatch between the ligature mark on the neck and the soft folded bedsheet recovered from the cell, do not cleanly support the suicidal hanging hypothesis that the medical examiner's office has maintained. Whether the alternative, ligature strangulation by a third party using an implement that was not the bedsheet, can be established to the standard of criminal proof is a separate question from whether the physical evidence is consistent with it. In my assessment, the physical evidence is more consistent with ligature strangulation than with suicidal hanging from the means described in the official scene reconstruction. That is a forensic assessment, not a criminal conclusion, and I am offering it as such.

The video record, as I have discussed above, has been substantially compromised by the editing and metadata anomalies documented by independent forensic analysts in 2025. The released footage does not cover the relevant cell door or several of the access routes to the tier, and therefore cannot definitively establish either that someone entered the cell during the relevant hours or that no one did. The government's public claim to the contrary, based on this footage, cannot be sustained on the basis of what the footage actually shows.

The fuller picture of Epstein's network of contacts and the extent to which powerful individuals had motives that might have been operative in a scenario of third-party involvement is being assembled slowly through the releases under the Transparency Act, through the House Oversight Committee documents, and through the ongoing civil litigation involving the estate. I do not claim, on the basis of the documents released to date, that any specific individual or group is implicated in Epstein's death. I claim only that the original physical evidence, combined with the subsequent compromise of the video record and the ongoing resistance of the executive branch to full disclosure, does not permit the conclusion that the case is closed, and that the forensic questions are resolved. They are not.

Summary

The death of Jeffrey Epstein, six years after it occurred, remains a case in which the physical evidence points in one direction and the institutional conclusion points in another, and the intervening years have not closed that gap but have widened it. The July 2025 Department of Justice and FBI memorandum declaring the case resolved was followed within weeks by forensic demonstrations that the supporting video evidence had been edited in ways inconsistent with the description the government had provided of the material. The footage released to the public does not show Epstein's cell door, does not show the main entry points to the unit, and therefore cannot substantiate the claim that no one entered the area where Epstein was held during the relevant period. The ligature mark on Epstein's neck is not consistent in width, texture, or residue pattern with the soft folded bedsheet recovered from his cell, and no alternative implement has been identified. The multiple neck fractures, the petechial hemorrhaging, and the absence of blood or skin transfer onto the recovered ligature are each individually unusual in suicidal hangings and are collectively more consistent with ligature strangulation than with the official cause of death. The institutional failures at the facility, the guards' behavior, the missing cellmate, the malfunctioning cameras outside the cell, were real, are documented, and created the window in which whatever happened could happen.

I owe a correction to my earlier article on the matter of the nasal tip morphology in the post-mortem photograph, which I now believe is more plausibly attributable to the degradation of hyaluronic acid fillers over time than to identity substitution, and the fingerprint identification of the body resolves the question of whether the corpse was Epstein. It was Epstein. The question is how Epstein came to be the corpse, and that question remains, in my forensic judgment, unresolved.

The Epstein Files Transparency Act, signed into law in November 2025, may in time produce a record sufficient to resolve some of what cannot currently be resolved. The initial December 2025 release did not. The department's continuing pattern of partial disclosure and extensive redaction, in substantive tension with the statute that Congress just passed, is not a pattern that supports confidence in the completeness of the record. Those who conclude that the official narrative is complete and settled are doing so on the basis of evidence that has been shown to be edited, incomplete, and in important respects inconsistent with the physical findings from the body. Those who conclude that the case cannot be considered closed are doing so on the basis of the same evidence, read more carefully. I place myself in the second category, and I do so not as a conspiracy theorist but as a forensic professional through and through, with over twenty-five years of practice, who has learned to see connections others overlook, and who knows that it is the physical evidence that tells the more reliable story.

The forensic record, read honestly, does not support the suicide verdict. It does not conclusively support any alternative verdict either. What it supports is the conclusion that the case was closed prematurely, on the basis of evidence that was incomplete at the time and has since been shown to have been compromised. Anyone who accepts the closure without asking why the evidence had to be edited, why the cell door was never on camera, why the ligature did not match the wound, and why the multiple fractures were consistent with the wrong mechanism, is accepting the closure for reasons that do not include the evidence.

This article is intended for general informational purposes and represents the author's analysis of publicly available evidence and scientific literature. Nothing in this article constitutes a formal forensic opinion issued in connection with any active criminal proceeding.

References

• Axios (2025, July 7). Exclusive: DOJ, FBI conclude Jeffrey Epstein had no client list, committed suicide.
• CBS News (2020, January 5). Jeffrey Epstein's autopsy: a closer look. 60 Minutes.
• CBS News (2025, July 28). Jeffrey Epstein jail video reveals new discrepancies.
• CBS News (2025, September 22). Jeffrey Epstein jail video investigation.
• CNBC (2025, December 24). Epstein files: DOJ says one million new documents found.
• Congress.gov (2025). H.R. 4405 Epstein Files Transparency Act, Public Law 119-38, November 19, 2025.
• Connecticut Examiner (2026, February 17). Epstein's autopsy deserves a second look.
• Daily Beast (2025, July 29). FBI has secret Epstein prison tape with no missing minute.
• DOJ/FBI Joint Memorandum (2025, July 7). Review of Epstein investigation files.
• DOJ Office of the Inspector General (2023). Report on the death of Jeffrey Epstein at the Metropolitan Correctional Center.
• Fox News (2025, September 2). Missing minute from Jeffrey Epstein jail security video revealed in document dump.
• Khokhlov, V.D. (2001). Calculation of tolerance limits when analyzing the occurrence of fractures of hyoid bone and laryngeal cartilages in hanging. Forensic Science International, 121(1-2), 32-34.
• Ligature mark on neck: how informative (2005). Journal of Indian Academy of Forensic Medicine, 27(1), 10-15.
• Master, M. (2024). Hyaluronic acid filler longevity in the mid-face: a review of 33 magnetic resonance imaging studies. Plastic and Reconstructive Surgery Global Open, 12(7).
• Wired (2025, July 15). Metadata analysis of DOJ-released Epstein surveillance file.

I will say, after six years of watching this case, and after reviewing what the past fifteen months of disclosure have added to what was already a troubling record, what my forensic judgment and my twenty-five years of practiced instinct together permit me to say plainly. Jeffrey Epstein did not hang himself in that cell. The physical evidence does not support it, the video record does not support it, and the sequence of institutional failures that created the window for his death supports it only if one accepts a probability so vanishingly small that no forensic investigator operating under an ordinary burden of proof would accept it. In Germany, this case would not have been closed. In any jurisdiction in which the men who stood to lose from Epstein's testimony did not also control the apparatus that investigated his death, this case would not have been closed. It was closed because it was closed by them, or with their indifference, and the decision served their interests in a way no forensic examiner can responsibly ignore. Epstein was murdered. I have said it with twenty-five years of forensic work behind me, and with the full understanding that in saying it I am stating a conclusion that the government of the United States has formally rejected. My intuition has not failed me in this field. It is not failing me now.

This article is the revised and expanded version of the original publication from 28 February 2025, updated in light of the 2025 document releases, the forensic analysis of the edited surveillance video, and subsequent scientific literature. Republished on 24 April 2026.