--- aliases: - Project Bluebird - Project Artichoke - ARTICHOKE - BLUEBIRD category: Intelligence Operation created: 2026-05-17 end: 1956-01-01 location: Washington, D.C. / overseas start: 1950-04-20 summary: Project Bluebird (1950-1951) and its successor Project Artichoke (1951-1956) were the CIA's classified behavioral research programs that pioneered LSD interrogation experiments on prisoners and foreign nationals, forming the direct institutional predecessor to Project MKULTRA. tags: - Program - CIA - MKULTRA - HumanExperimentation - Interrogation - ColdWar - 1950s title: Project Bluebird and Project Artichoke updated: 2026-05-17 --- Project Bluebird (authorized April 20, 1950) and its successor Project Artichoke (authorized 1951) were the [CIA](/organizations/central-intelligence-agency/)'s first systematic behavioral research and interrogation programs, established to develop methods for controlling human behavior, inducing reliable amnesia, and extracting information from resistant subjects through pharmacological and psychological means. Together they represent the direct institutional predecessor to [MKULTRA](/programs/project-mkultra/), which absorbed and expanded their mandate in 1953 under [Sidney Gottlieb](/people/sidney-gottlieb/).[^1] ### Project Bluebird CIA Director Roscoe Hillenkoetter authorized Project Bluebird on April 20, 1950, in response to concern about Soviet and Chinese interrogation techniques apparently used on American prisoners of war during the [Korean War](/events/korean-war/). The program's authorization memorandum defined three objectives: 1. Whether reliable amnesia could be induced in individuals who performed clandestine operations, to prevent capture-and-interrogation from revealing operational details 2. Whether foreign agents could be "reconditioned" through a combination of hypnosis, drugs, and psychological manipulation 3. Whether methods could be developed for controlling the behavior of individuals without their awareness or consent Bluebird was managed by the CIA's Security Research Service and was primarily focused on hypnosis and narcosynthesis (drug-assisted hypnosis) rather than the broader pharmacological research that characterized later programs. Its subjects were principally foreign prisoners in overseas CIA facilities, where legal constraints on experimentation were less immediately applicable.[^1] ### Project Artichoke Project Artichoke superseded Bluebird in 1951 under the oversight of the CIA's Office of Scientific Intelligence. Artichoke expanded the program's mandate and its toolkit: [LSD](/concepts/lsd/) (recently synthesized by Sandoz Pharmaceuticals) was added to the pharmacological arsenal for the first time in a CIA context, alongside mescaline, sodium pentothal, barbiturates, and other psychoactive agents. Artichoke's operational scope was also broader. The program conducted experiments on: - Foreign prisoners held in CIA overseas facilities in [Japan](/places/japan/), [Germany](/places/germany/), and elsewhere - Korean War POWs (in medical examination contexts) - CIA defectors undergoing security vetting - In at least one documented case, an unwitting American subject in a domestic setting - a case that the [Church Committee](/events/church-committee/) subsequently identified as an early instance of the pattern of domestic experimentation that would characterize MKULTRA Artichoke's central operational question - whether a subject could be conditioned to perform an action against their own interests (including, in one documented case study, assassination) without any subsequent memory of having done so - became a recurring preoccupation of the program's internal documentation.[^1] ### Sheffield Edwards and the First Teams Sheffield Edwards, the CIA's Security chief, launched Bluebird in April 1950 by convening all interested Agency parties and proposing dedicated interrogation teams, each composed of a psychiatrist, a polygraph expert trained in hypnosis, and a technician. Operational research was placed under [Morse Allen](/people/morse-allen/), a security operator who had spent his career pursuing domestic communists. Hillenkoetter authorized unvouchered funds, and early documents stated the goal of "controlling an individual to the point where he will do our bidding against his will and even against such fundamental laws of nature as self-preservation"; one officer warned, "If this is supposed to be covered up as a defensive feasibility study, it's pretty damn transparent."[^3] ### Field Operations and Operation CASTIGATE Three months after approval, Allen led the first team to Japan in July 1950, a month into the Korean War, testing combinations of sodium amytal and benzedrine on four suspected double agents and attempting to induce amnesia; around October 1950 the team applied "advanced" techniques to 25 subjects described as North Korean prisoners of war. In 1952 Allen led the Artichoke team to Frankfurt, Germany for [Operation CASTIGATE](/programs/operation-castigate/), where [G. Richard Wendt](/people/g-richard-wendt/)'s truth-drug experiments failed; Allen called the trip "a waste of time and money" and recommended that drug developers be barred from field testing and, "as an absolute rule," that no women accompany Artichoke missions.[^3] ### Terminal Experiments and Interagency Liaison Allen and his coworkers developed the concept of "terminal experiments," tests carried through to completion on subjects who had no idea they were involved, on the principle that only subjects "for whom much is at stake (perhaps life and death)" yielded operationally meaningful results. [Harold Wolff](/people/harold-wolff/) of Cornell Medical School, proposing harmful experiments, wrote that "we expect the Agency to make available suitable subjects and a proper place for the performance of the necessary experiments." In April 1951 the Director approved liaison with Army, Navy, and Air Force intelligence; the [FBI](/organizations/federal-bureau-of-investigation/) declined to participate, and at the first trilateral meeting the British representative remarked that there had been nothing new in interrogation since the Inquisition.[^3] ### The Artichoke Assassination Question One of the more disturbing elements of the Artichoke documentation that survived is a January 1954 memorandum discussing whether the program had succeeded in creating a "Manchurian Candidate" - an individual programmed to carry out a violent act without conscious volition or subsequent memory. The memorandum describes experimental sessions in which subjects were placed in hypnotic trances and instructed to perform tasks they would normally refuse; the document discusses whether similar conditioning could be applied to assassination. Whether any operational assassination was actually carried out through Artichoke conditioning is not established in the surviving record. The memorandum describes the theoretical question rather than a completed operation. Journalist John Marks, who discovered the surviving MKULTRA documents through FOIA, considered this memorandum significant enough to structure his investigation around the question it raised.[^2] ### Morse Allen's Hypnosis Experiments On February 19, 1954, Allen ran a now-notorious test on Agency secretaries working late: he hypnotized one into a deep sleep and told a second that her rage at being unable to wake her colleague would be so great "that she would not hesitate to 'kill,'" leaving an unloaded pistol within reach. The second secretary, who had previously expressed a fear of firearms, picked up the gun, "shot" her sleeping colleague, and afterward had apparent amnesia. Allen concluded only that "an impressionable young volunteer would accept a command from a legitimate authority figure" she trusted would prevent real harm, and remained unconvinced that hypnotic programming would work on resistant or unwitting subjects. A January 22, 1954 memorandum framed the question explicitly, asking whether "an individual of [redacted] descent [could] be made to perform an act of attempted assassination involuntarily under the influence of ARTICHOKE," proposing that a target be "surreptitiously drugged through the medium of an alcoholic cocktail at a social party," hypnotized, and programmed; handwritten notes marked the assassination component "simulated only." Through 1954 Allen lobbied to run true terminal hypnosis experiments on a foreign agent, a scheme that reached advanced planning, with messages exchanged about the availability of "time, place, and bodies," before a station chief's second thoughts ended it. By the end of 1954 Dulles transferred the behavioral function from Allen to Gottlieb and MKULTRA.[^3] ### Camp King and the Nazi Connection While Bluebird operated, the CIA and Army ran an overlapping interrogation program at [Camp King](/places/camp-king/) in occupied Germany, whose chief medical officer was the former Nazi surgeon general Walter Schreiber, brought over under [Operation Paperclip](/programs/operation-paperclip/); when Schreiber was exposed and moved to the United States, the post passed to Kurt Blome, who had directed the Nazi biological-weapons program. The Paperclip scientists there contributed to Bluebird's approach of combining pharmacological agents with psychological pressure, and the Camp King files remain among the most heavily redacted in the CIA's CREST database.[^4] ### Transition to MKULTRA In 1953, CIA Director [Allen Dulles](/people/allen-dulles/) authorized Project MKUltra on the basis of a proposal from Sidney Gottlieb that explicitly built on the Artichoke and Bluebird foundations while expanding them substantially. MKULTRA absorbed the institutional knowledge, some of the personnel, and the operational framework developed under Artichoke, while scaling the program through the cut-out foundation structure that allowed much broader research engagement. The formal distinction between the programs is primarily organizational: Bluebird and Artichoke were run through the Office of Scientific Intelligence; MKULTRA was run through the Technical Services Staff and later Technical Services Division. The subject matter, methods, and ethical failures were continuous across all three programs. Artichoke continued nominally as a separate program designation for some time after MKULTRA's authorization, handling certain overseas interrogation operations that the CIA maintained as formally separate from MKULTRA. By the late 1950s the distinction was primarily administrative. ### Church Committee Findings The Church Committee's 1976 investigation examined Bluebird and Artichoke as part of its review of CIA domestic and foreign intelligence activities. The committee found that both programs had used unconsenting subjects, had conducted experiments abroad in jurisdictions where American legal protections did not apply, and had shared the fundamental ethical failure that characterized MKULTRA: the absence of any institutional mechanism for weighing subject welfare against research goals.[^1] [^1]: Church Committee (U.S. Senate Select Committee to Study Governmental Operations with Respect to Intelligence Activities). Final Report, Book I ("Foreign and Military Intelligence"), S. Rept. 94-755, April 26, 1976. Marks, John. *The Search for the Manchurian Candidate: The CIA and Mind Control.* Times Books, 1979, Chapters 2-3 (primary account of Bluebird and Artichoke using surviving documents). [^2]: CIA memorandum, "Artichoke," January 1954. Reproduced in part in Marks, *Manchurian Candidate*, pp. 27-28. CIA CREST database, FOIA-released documents, accessible at cia.gov/readingroom. [^3]: John D. Marks, *The Search for the Manchurian Candidate*. Times Books, 1979, Chapters 2, 3, and 11; National Security Archive, "Document Friday: Project ARTICHOKE," April 23, 2010; Alliance for Human Research Protection, "1954: Can an individual be made to perform assassination involuntarily under the influence of ARTICHOKE?" [^4]: "What Cold War CIA Interrogators Learned from the Nazis," *The Daily Beast*, 2014; National Security Archive, "The CIA and Nazi War Criminals," NSAEBB 146.