MKULTRA

The record and its limits

What the existing documents establish.

Original records establish funded research through universities, hospitals, contractors and intermediaries, including unwitting human tests. Most operational files were destroyed in 1973; financial files survived.

Accountability / research target

The question the original records still need to answer.

Reconcile the 149 project-number slots with original vouchers, intermediaries, experiment categories, actual recipients and people affected.

This dossier will expand as original agency authorizations, financial records and contemporaneous correspondence are matched to named events. An unresolved request remains an investigation target, not a finding about an individual.

Direct documentary sources

Read the government and archival evidence.

Internal source register IDs: MK-002, MK-010, MK-024. Open the complete source register ↗

R12 / direct CIA briefing book page review

Seven newly extracted subproject records: 8–14.

These separate entries are research and administrative instruments, not seven confirmed instances of experimental harm. Source links point to the exact released CIA briefing-book pages, and the estimated project figures are not combined into an aggregate without checking interrelated contracts and overlaps.

Twenty-six of the 149 numbered entries now have source-linked preliminary descriptions. The other 123 remain unreviewed at page level. Open source-graded 149-slot workbook ↗

R15 / preserved CIA documents in the 1977 Senate hearing

The payment, the hospital and the lawyer's memorandum.

Two contemporaneous documents preserved in the Senate record allow separate accountability questions to be pursued. The first records chemical testing on hospital patients with terminal cancer and hospital bills paid to facilitate that work. The second records CIA legal advice about routing agency funds through a private-appearing gift and how that arrangement might affect government matching funds and hospital overhead. Their financial purposes and beneficiaries differ and cannot be combined into a single transaction.

The same Senate record reproduces a Subproject 16 facility-support memorandum estimating $7,740 for one year's maintenance of sites established under Subproject 3. This is a budget estimate, not evidence that every contemplated test took place. Read the source page. These findings are linked in the 149-slot workbook and evidence register. Relevant page text was checked, although the archive screenshot service did not return full images for this review; the unresolved image-verification limitation is preserved in the research checkpoint.

Source-graded program genealogy

Related program records.

• MKSEARCHportion continued under

  explicit

• QKHILLTOPabsorbed work

  explicit

• Safehouse testing / Operation Midnight Climaxincluded covert testing operation

  historically documented

Relationship labels distinguish renamed or directly absorbed projects from subject overlap and later historical comparison. Every connection can be corrected if a contemporaneous record conflicts with the present classification.

Separate original-file cross-reference

Connection to the Epstein investigation.

No operational connection between this specific historical program and Epstein or an Epstein associate is established in the currently registered primary evidence. A possible match can be assessed in the companion Epstein files investigation using exact original file identifiers and matching institutional dates. A later research interest, shared subject term, or geographic overlap is insufficient.

R14 / authorization, expenditures and cancelled work

The amounts in the CIA record are different kinds of evidence.

A procurement ceiling does not establish payment, and a retrospective figure can appear even when an experiment did not occur. Select a subproject to inspect how the briefing describes the work and the money. These entries are drawn from one original CIA compilation and cannot be added into a meaningful expenditure total.

• Subproject 17: approximately $30,000 via Geschickter Foundation, with no underlying payment voucher reviewed. Original page 67.
• Subproject 18: $400,000 authorized, $1,337.45 expended. Original page 70.
• Subproject 20: approximately $202 associated with chemical supply. Original page 74.
• Subproject 21: a $5,800 approximate-total field alongside an explicit zero-expenditure statement. Original page 76.

Forty-two of the 149 numbered slots now have source-page summaries. The other 107 remain pending. Draft records-request docket RD-18 and RD-19 specifies the original finance and project-closure records needed to test this accounting.

Document-to-document relationships

One subproject number does not always mean one separate experiment or payment.

Several entries in the CIA briefing identify follow-on work, an accounting correction or operational support. The links below distinguish those categories from entries in which the agency itself described human drug testing. The record is still incomplete: figures come from a retrospective briefing, and consent documents and underlying payment vouchers have not been independently authenticated.

This comparison includes source-text abstracts for Subprojects 22, 24–26 and 28–33. Cases 16, 23 and 35 retain separately documented archival exhibits. Entry 26 refers back to 9, entry 32 continues 12, and entry 33 is a $400 adjustment to 27. The named institutions and researchers should be assessed through original signed grants and ethics records before attributing what they knew.

Decision and accountability examiner

What would establish what actually happened?

The released CIA descriptions carry different evidentiary weight. Testing is explicitly described for some projects; other entries document proposed work, cover funding or an internal refusal to approve money. Choose a record below to distinguish what the cited source establishes from the additional instruments needed to verify consent, institutional awareness and actual expenditures.

These comparisons have no historical subject identifiers and do not assert that any Epstein-connected institution or person participated. The original-source request docket contains narrowly defined institutional inquiries. Some briefing pages were verified through machine-extracted page text because the archival image server intermittently failed; their evidence status is recorded in the workbook.

Institutional funding routes / source inspector

Who provided the money, and who could see its origin?

The CIA's surviving retrospective brief identifies financial intermediaries that concealed federal sponsorship of scientific and behavioral research. Its amounts combine estimates, later authorizations and different project reporting windows. The records can establish an authorized routing structure without establishing that every institution knew the real donor or that every reported dollar was disbursed.

Draft accountability requests RD-27 to RD-29 seek the original intermediary ledgers, transfer authorities, institutional awareness records and de-identified experimental oversight materials. No request has been submitted, and these historical routing diagrams establish no later operational connection to the Epstein investigation.

First-party document comparison

Which records would settle the unresolved questions?

Historical research summaries, signed financial authorizations, payments and informed-consent evidence occupy different documentary categories. This inspection table compares each category without treating the absence of surviving records as proof of a particular outcome. The original CIA record for Subproject 68 includes an unresolved patient-awareness issue, and the September 1977 hearing distinguishes the MKSEARCH financial folders from twelve related but separate grant files.

Of the 149 numbered subprojects, 71 now have preliminary source-linked summaries and 78 remain for page-specific review. PDF extraction and page images are recorded separately in the source workbook. The above does not establish any historical Epstein operational connection.

Finance and institutional accountability

Why the archived spending figures cannot be totaled yet

The CIA's original 149-number briefing uses approximate figures for different accounting scopes. The Subproject 74 total refers to a portfolio with more than 25 recipients, while a separate, indexed CIA award contains an unaccounted balance. Other entries contain smaller charges within larger projects, or describe intermediary payments without revealing actual downstream amounts. This investigation keeps those categories separate until independent grant instruments and paid vouchers support reconciliation.

All 149 numbered slots have preliminary CIA-source summaries. Some individual descriptions were extracted from scanned or indexed text when original signed grant images were inaccessible. Retrospective summary amounts cannot be treated as verified expenditures without matching authorization and payment records. No MKULTRA-to-Epstein operational connection has been established in the reviewed originals.

Institutional responses / competing original records

The CIA's original summary also records institutional corrections

The same surviving intelligence file can contain a description of Agency interest and a university or organization's correction of that description. The historical record requires readers to keep both. An intention to study or recruit does not establish that an intervention occurred. A grant routed through a research society does not establish that a university controlled the money or knew the ultimate sponsor.

All 149 numbered subprojects now have preliminary CIA-source summaries. Original signed awards, individual trial protocols and grant-level payment and notification documents remain only partially recovered. Clinical, youth and refugee participant identities remain protected. The existing Epstein-files research is separate; no operational connection has been established.

Original-source reconciliation

The historical program was not 149 separate experiments

CIA administrative records link Subprojects 42, 132 and 149 and later MKSEARCH accounting. They also distinguish policy pauses, rent, program budgets and actual experiments. A numbered renewal or rental charge cannot be treated as another verified human drug test. The original CIA report includes a late testing stand-down and the agency's later assessment of whether follow-up was necessary. The historical documents have different reliability limits and some PDF page images are not currently accessible.

Financial distinction: the original Subproject 133 draft continuation memorandum separately records $9,000 proposed direct work and an included $360 four-percent administrative fee, totaling $9,360. The original paid-voucher record remains open. Three archival-request drafts RD-55–RD-57 remain unsent. No historical MKULTRA-to-Epstein operational connection is established.

Financial reconstruction

One funding transfer, three accounting levels

The original CIA fiscal collection shows a $74,000 reassignment in 1956, separate from the later fiscal-year $52,150 grant obligations. An internal reallocation is not a new independent paid award.

CIA original Subproject 65 financial file (31 pages, index text reviewed). Signed payments and complete original page images have not been authenticated. RD-64 remains unsent.

Original inquiry unanswered

What CIA investigators still sought in 1979

An April 1979 CIA letter asked Sidney Gottlieb to help explain incomplete records, the health follow-up of research participants, and the distinction between original locations and later accounts of operations. A question asked by investigators is not proof that an event occurred or did not occur.

Source: original CIA letter, National Security Archive. The remaining archival requests RD-65 and RD-66 are unsent.