35+ Years of Freedom of Information Action

CIA declassifies more of "Zendebad, Shah!" – internal study of 1953 Iran coup

Shah Mohammed Reza Pahlavi of Iran and President Dwight D. Eisenhower – all smiles in the early years following the ouster of Prime Minister Mohammad Mosaddeq. (RFE web site)

Published: Feb 12, 2018
Briefing Book #618

Edited by Danielle Siegel and Malcolm Byrne

For further information, contact: Malcolm Byrne, 202/994–7000 or nsarchiv@gwu.edu

Dozens of formerly Secret passages released, but many operational details still withheld

Internal critique of tensions between CIA directorates sheds light on U.S. approach to covert activities

Washington, D.C., February 12, 2018 – A partially-declassified CIA history of the 1953 coup in Iran, released in late 2017, includes an-depth critique of how the agency approached the operation, highlighting the effects of bureaucracy and politics on the conduct of U.S. clandestine activities. The CIA report, posted today by the George Washington University-based National Security Archive, also reveals details about the hatching of the covert plot as well as its execution. 

Portions of the history, “Zendebad, Shah!”, have been made public twice before – in 2000 and 2011 – but those versions were identical and provided only the unclassified passages of the roughly 139-page document, withholding the rest on secrecy grounds. Last year, in response to the most recent Mandatory Declassification Review request by the National Security Archive, the CIA conducted another re-review, this time producing considerably more text, although still holding back a substantial amount. 

Today’s posting includes side-by-side examples to illustrate some of the differences in releases over the years.


Differences between 2011 and 2017 releases

The latest version adds 37 pages of previously unreleased appendixes in addition to filling in many excised passages throughout the rest of the text. It also helpfully provides most of the Table of Contents – previously (and inexplicably) withheld in full. Despite the expanded access, considerable information on planning and carrying out the coup is still being denied on classification grounds. In addition, two of five appendixes – including their titles – have been withheld. The Archive is appealing these continued deletions.

One likely explanation for the latest, expanded release of "Zendebad, Shah!" is the fact that in June 2017 the State Department finally authorized publication of a long-awaited volume on the coup from its Foreign Relations of the United States series. That compilation contained a substantial amount of documentation that had been kept from public access for over 60 years. CIA declassification review officials presumably based many of their decisions on the same criteria that governed release of the FRUS materials.


Why are there three CIA histories of TPAJAX?

Since 1953, the CIA has produced three separate internal accounts of the coup. The earliest was written by one of the plan’s architects, Donald Wilber, in March 1954, a few months after the fact. It is by far the most detailed about the planning and rollout of the operation and clearly aimed at providing lessons for future covert operatives. But according to current CIA officials it had shortcomings. One was the author’s subjectivity. In the words of one official, “Wilber had axes to grind,” primarily with Kermit Roosevelt, the CIA’s operational leader, who “got all the attention.”(i) Another reason was a lack of interviews with other coup participants; Wilber relied largely on his own recollections – in addition to documentation available to him (although the FRUS volume shows that he was able to get Roosevelt’s reactions immediately after the operation). Finally, his account was entirely about the coup’s operational aspects, with no discussion or assessment of the analytical side of the process.

The Wilber history remains classified despite a 1999 Freedom of Information Act lawsuit by the National Security Archive. But The New York Times obtained a leaked copy and posted the bulk of it minus certain names and identifying details on its web site in 2000. That version is also available here.

The second history, The Battle for Iran, is undated but appeared sometime in the mid-1970s. Judging from its focus, tone, and the time period in which it was written, one of the author’s goals was clearly to defend the agency’s actions. The CIA at the time was the target of public opprobrium after public revelations of domestic spying and other illegalities around the world. Implicitly countering the frequent charge that it acted like a “rogue elephant,” the author makes a point of insisting that TPAJAX, the codename for the 1953 coup, was “conceived and approved at the highest levels of government” [p. 26]. A surprisingly large part of the document is then devoted to reproducing – and criticizing – excerpts from media and other published accounts that purportedly misrepresent facts about the coup or the agency.

The last known internal CIA history, written in 1998, was “Zendebad, Shah!” The Central Intelligence Agency and the Fall of Iranian Prime Minister Mohammed Mossadeq, August 1953. According to CIA officials, one of the aims of this version was to describe the analytical side of the agency’s actions during the coup. Also, there was a general view that earlier problems with access to available documentation had been resolved, so a qualified historian on the agency’s staff would finally be in a position to write the fullest possible account (although the author’s prior level of expertise on Iran is unknown). Mohammed Khatami’s surprise 1997 election as Iran’s president – a development of keen interest to President Bill Clinton – may also have contributed to renewed interest in the coup.


Dubious Secrets

With one exception [p. 39], earlier public releases of "Zendebad, Shah!" in response to the National Security Archive’s FOIA requests featured only the paragraphs the agency had originally marked Unclassified. The current release makes available a considerable number of formerly Secret passages. As indicated, many of them track with what was published in the retrospective FRUS volume, although there are also a few differences.

Every occurrence of the word “Shah” was assiduously redacted from the 2011 declassification review.

In a number of cases, what had previously been withheld was trivial or otherwise arguably unjustified. Unflattering references to Mosaddeq, for instance, were whited out in an earlier version [p. 4]. This might be seen as laudable except for the irony that U.S. officials in the lead-up to TPAJAX publicly and privately disparaged him, sometimes in the crudest terms. 

Other passages were excised in "Zendebad, Shah!" that had never been classified in the first place. An example is the text of one of the Shah’s firmans (orders) relating to Mosaddeq’s ouster – another irony given that one of the keys to the operation’s eventual success was the decision to disseminate those directives far and wide after the first attempt collapsed. 

Other dubious deletions include footnotes citing unclassified documents, and even excerpts from Roosevelt’s published memoir, Countercoup. (An earlier release of The Battle for Iran likewise whited out whole passages from mass market books published decades ago.)

Misguided classification may not be entirely avoidable, especially where there’s a lack of familiarity with the subject. But a larger problem of grossly excessive classification has hampered public understanding of the 1953 coup. The great bulk of the history could and should have been publicly released many years ago, and undoubtedly would have been except for the persistent refusal of segments of the intelligence and diplomatic communities who, it turns out, were responding to a certain degree to the requests of the British government. In recent years, U.S. official reluctance began to fade although, in a final irony, politics likely played a part in the FRUS release in June 2017. (Visit the Archive’s Iran-U.S. Relations web page to see previous Electronic Briefing Books discussing these issues.) But the bottom line is that U.S. interests and credibility arguably suffered as a result of a persistent refusal to acknowledge the reality that the coup had long since ceased to be a secret.


Items of Note in the latest "Zendebad, Shah!" Version

Followers of the coup saga will find a variety of new – though not necessarily revelatory – detail in today’s posted document. Some of it tracks to a degree with what appeared in the June 2017 FRUS volume. But while the FRUS series by design takes no positions on the topic being covered, "Zendebad, Shah!" is a piece of analysis by a historian who includes his own viewpoints on various issues. His standpoint on the relative importance of the CIA and Iranian actors in the coup’s success, and especially his take on Roosevelt’s role, will probably generate pointed reactions from some readers.

By far the biggest contribution of the new "Zendebad, Shah!" version is its detailed discussion of the interaction (or lack thereof) between the CIA’s operational and analytical divisions. That will interest intelligence experts as much as Iran watchers. 

Here are a few of the areas of interest in the latest release:



A key missing piece of the documentary record on the coup is President Eisenhower’s authorization. The Wilber history gives a specific date (July 11, 1953) and the FRUS volume notes a handwritten mark on a document indicating the same. But the FRUS compiler adds: “No written authorization from the President has been found.” "Zendebad, Shah!" is not much help. On page 18 it says that Eisenhower “apparently” made the decision. On page 20, it reads: “Available documents do not indicate who authorized CIA to begin planning the operation, but it almost certainly was President Eisenhower himself.” Tellingly, the author supports this statement not with an internal document but with a quote from Eisenhower biographer Stephen Ambrose.



Mosaddeq arriving for his trial in Tehran in November 1953. (Iran Chamber Society)

The 2017 version of "Zendebad, Shah!" offers new insights and details surrounding the planning, execution, and follow-up stages of TPAJAX. First, it notes that CIA officials thought that a Zahedi government would be strong enough in the short-term to serve as an alternative to Mosaddeq, but would struggle to survive in the long-term because “The Shah had never supported any of his past Prime Ministers consistently, and CIA did not believe this would change.” [p. 70] This assessment of Zahedi as a short-term solution is particularly interesting in light of how the CIA characterized what was at stake in Iran. Agency operatives saw political chaos and instability as catalysts for empowering the Communist Tudeh party, and thought the best bulwark against Soviet subversion was a strong, stable, and competent central government [p. 11]. This new information illustrates that the CIA actively sought to replace Mosaddeq with a leader whom they knew would struggle to garner a long-term base of national support, and whom they expected would perpetuate the problems of division and discord in Iranian politics.

The latest version of "Zendebad, Shah!" also provides an interesting take on the fundamental shift in approach that occurred between August 15th and August 19th. TPAJAX, the author writes, was originally designed as a military coup “with the aim of installing Zahedi as Chief of the General Staff and then Prime Minister” [p. 48]. This attempt failed on August 15th when General Nassiri was arrested while delivering the firmans signed by the Shah appointing Zahedi as the new prime minister. Because the arrest of Nassiri led to the disintegration of Zahedi’s support within the army and the retreat of his troops, any subsequent attempts to oust Mosaddeq could not rely on military backing. The signed firmans, the report notes, “changed the entire character of TPAJAX” because the end goal was no longer to execute a military coup, but to convince the public that the Shah was exercising his constitutional authority and Mosaddeq’s refusal to acquiesce “would make his tenure as Prime Minister illegal” [p. 48]. Thus, the second and final attempt to oust Mosaddeq on August 19th focused on coalescing public support for the Shah and, the author contends, is more accurately characterized as “a rapid, concentrated, and effectively improvised psychological warfare campaign” than a carefully-planned military coup [p. 55].

General Nassiri getting his general’s stars for his role in helping to oust Mosaddeq. (The Iranian)

Finally, this new version of "Zendebad, Shah!" offers some discussion of Mosaddeq’s trial, including the Shah’s attitudes and the CIA’s role in shaping how both the Zahedi Government and the Shah handled the proceeding and its outcome—although a key detail appears to have been excised. The report cites a CIA cable urging the Shah not to allow a public trial, despite his inclination to do so, to “avoid any action that might make Mossadeq a martyr” [p. 110]. It goes on to conclude that “considerable pressure” from Washington was likely the main reason why the Shah ultimately resisted his initial desire to make the trial public. Furthermore, the same CIA cable advised Zahedi to keep Mosaddeq in “’enforced residence in a small village’ under strong surveillance” [p. 110].

Some of these details, and more, appear in the new FRUS volume. What is missing in "Zendebad, Shah!" is the fact that Zahedi wanted the immediate execution of Mosaddeq, and the Shah, too, was ready to executive him if the Tudeh threatened to take any action after the trial. (This may be what is currently excised on pages 110 or 112).



The 2017 release of "Zendebad, Shah!" also adds to the ongoing historical debate about whether the CIA can take responsibility for the fall of Mosaddeq or if his overthrow was a consequence of internal forces within Iranian society. The report contains some new bits of information recognizing the internal Iranian factors at play, like the importance of mullahs in convincing the people of South Tehran that Tudeh protests were an act of Soviet subversion allowed by Mosaddeq [p. 63]. However, it overwhelmingly concludes that without CIA direction, a Mosaddeq coup would not have been possible. The author asserts that after the failed August 15th attempt and General Nassiri’s arrest, “Riahi’s and Mosaddeq’s quick reactions effectively emasculated the military’s participation in TPAJAX” [p. 50]. The author reports further, without citation, that “the principal anti-Mossadeq figures lost their courage” [p. 50], leaving it to Roosevelt to pick up the pieces [p. 55ff.].

General Fazlollah Zahedi, Mosaddeq’s successor after TPAJAX.  The author of "Zendebad, Shah!" suggests that without CIA backing, Zahedi would not have been able to garner enough public and military support to overthrow the government – a point of contention even today among those who debate who was ultimately behind the fall of Mosaddeq. (Wikipedia)

Perhaps more importantly, the report concludes that without the CIA, there would have been no viable candidates to replace Mosaddeq. “No potential prime minister was strong enough to command a majority in the Majlis, or even to form a coalition government out of the factions and splinter groups comprising Iranian politics” [p. 88]. In the author’s view, it was ultimately Kermit Roosevelt’s campaign to garner public support for the Shah and back Zahedi that created an alternative to Mosaddeq in the absence of an obvious successor.

In keeping with other recently available evidence – and, for that matter, information that has been known for decades – the newly-released passages of "Zendebad, Shah!" will likely generate more debate over who made the coup finally succeed. But readers will also experience a certain déjà vu feeling since "Zendebad, Shah!’s" author largely made use of the same sources that have finally started to come out in FRUS or from Freedom of Information Act releases in recent years.



Kermit Roosevelt, grandson of President Theodore Roosevelt, was the CIA’s operational point-person in Tehran during TPAJAX.  The author of "Zendebad, Shah!" asserts that he played an indispensable role in Mosaddeq's overthrow, even after the coup’s initial failure. (Harvard Magazine)

Even more likely to spark heated reactions are the document’s conclusions on the extent to which Kermit Roosevelt’s individual actions influenced the outcome of TPAJAX. On this question, "Zendebad, Shah!" argues in no unclear terms that Roosevelt was essentially the only reason the operation survived after the initial failure. As far as Washington and the Embassy in Tehran were concerned, TPAJAX was over on August 15th. This version of "Zendebad, Shah!" suggests that it is “no exaggeration” to say that the operation only succeeded because of Roosevelt and his quick decision to change the “essence” of TPAJAX “into a political action” when the military strategy proved no longer viable [p. 78].

At the same time, the author possibly unwittingly acknowledges the complexity of the situation with observations such as this: “it must be remembered that CIA was able to influence directly events only in the capital city, and there only barely. Kermit Roosevelt had neither the money nor the agents to initiate the kind of demonstrations that took place in Iran’s widely separated cities” [p. 86].

This version’s inclusion of Roosevelt’s comment to Eisenhower, Dulles, and the Joint Chiefs that he “made a point” of keeping the administration in the dark further illustrates just how removed the operation was from decision-makers in Washington during those fateful days between August 15th and 19th.



Anthony Eden, British foreign minister in 1953, may have played a key role in persuading certain U.S. officials, including Eisenhower, that Mosaddeq should be overthrown. (Wikipedia)

This version of "Zendebad, Shah!" also includes previously-classified sections detailing meetings between Washington and the British. It highlights a meeting between C.M. Woodhouse, M16 Chief of Station in Tehran, and CIA and State Department officials in November 1952 (right after Eisenhower’s election victory) during which Woodhouse stressed the vulnerability of Iran to Communist subversion and “did not stress the oil issue” – a strategy which resulted in Allen Dulles showing “increasing interest” in the idea of a covert operation to oust Mosaddeq [p. 15].

A few details of another previously-classified and extremely significant meeting between British Foreign Minister Eden and State Department officials in March 1953 are also included. During this meeting, “The State Department agreed with Eden that Mossadeq had to go, but its reasons differed from his. For Eden and his government, Mossadeq’s policies damaged ‘British prestige, influence, and vital commercial interest.’ For the Americans, Mossadeq represented a weakened Iran and its increasing vulnerability to Soviet domination’” [p. 16].

President Harry Truman and Prime Minister Mosaddeq during happier times – in Washington DC, October 1951, before Truman’s successor decided Mosaddeq was the problem not the solution to the Iran oil crisis.

These newly-released details in "Zendebad, Shah!" corroborate the findings of a document posted by The National Security Archive in August 2017 detailing a December 1952 meeting between British diplomats and State Department officials, when the British proposed American assistance in a coup to oust Mosaddeq. The memorandum of the conversation that occurred during this meeting illustrates how the British strategically framed what was at stake in Iran in terms of the Communist threat after previous attempts to persuade the Americans to get involved on the basis of protecting the Anglo-Iranian Oil Company had failed.



Perhaps the most significant difference between the 2017 version of "Zendebad, Shah!" and previous releases of the report is the declassification of an appendix exploring “The Tension Between Analysis and Operations” in the CIA during TPAJAX. This appendix paints an arresting picture of a divided CIA in which analysts were kept in the dark about the day-to-day details of the operation. It concludes that there is “no evidence” that “Kermit Roosevelt’s NEA Division consulted either the Office of National Estimates (ONE) or the analysts in CIA’s Office of Current Intelligence (OCI) at any stage of the operation” [p. 113]. Additionally, the Office of Strategic Operations often declined requests from ONE for updated intelligence [p. 114]. In other words, the analysts responsible for estimating the long-term strategic implications of TPAJAX and possible outcomes in Iran were doing so without updated and accurate knowledge of what was happening on the ground.

A major reason why this information gap between analysts and operatives was noteworthy, according to this version of "Zendebad, Shah!", is that it was the analysis wing of the Agency that was “responsible for keeping the President informed about daily events that might affect US foreign policy” [p. 116]. Thus, at least some of the intelligence briefings about U.S. operations the president was receiving were initially conducted “in ignorance of the developing American role” in Iran [p. 116].

Finally, this new appendix reveals that the tensions between analysts and operatives were not unknown to high-level agency officials. It concludes that “The highest levels of management in CIA did nothing to discourage estrangement of the Directorate of Plans from the Directorate of Intelligence, and in fact reinforced it. Allen Dulles ignored the Agency’s analytical arm during TPAJAX, preferring to use personal acquaintances as sources of information” [p. 118]. For example, the author notes that Max Thornburg, an oil company executive and consultant “for Middle Eastern governments on oil and economic questions” was one of Dulles’s primary sources for information about Iran [p. 119]. It further states that Thornburg “gained unusual access” to Dulles and “key State Department officials” [p. 119].



Finally, this version of "Zendebad, Shah!" includes the author’s perspective on how the legacy of TPAJAX influenced the standing of the CIA, the agenda of the Eisenhower administration, and the perceived efficacy of covert action as a tool of U.S. foreign policy. He notes that the CIA “did not conduct an investigation or post-mortem analysis of the operation to determine what the Agency did wrong and what it did right,” and subsequently argues that some of the most important lessons of TPAJAX were those that were made known but went unlearned [p. 79].

For example, the report cites a September 1953 meeting during which Roosevelt told Eisenhower, Dulles, the Cabinet, and the Joint Chiefs that the success of TPAJAX did not mean that the U.S. can “walk on water, everywhere” and urged exercising restraint before pursuing covert actions in the future [p. 79]. Yet, according to the author, neither the CIA nor the Eisenhower administration heeded Roosevelt’s warning and went ahead with a similar operation just a year later in Guatemala, called PBSUCCESS, based on “dubious conclusions” drawn from TPAJAX [p. 89]. He argues that “The rapidity and ease with which TPAJAX and PBSUCCESS had accomplished their objectives deceived CIA officials. They drew the erroneous lesson that the Agency could alter world events in the Third World at will and with minimal expense,” a lesson that would not be corrected until after the Bay of Pigs invasion in 1961, when the Kennedy administration finally undertook “a sober reappraisal of the utility and flexibility of covert action as a foreign policy instrument” [p. 81, 89-90].

Consequently, according to Koch, “Perhaps the most important and unique result of TPAJAX was to strengthen the CIA’s position within the government as an instrument of policy and solidify its responsibility for clandestine activity” [p. 90]. After the perceived successes in Iran and Guatemala, “any doubts Dulles may have had about TPAJAX and PBSUCCESS vanished as both operations unfolded and benefited from unexpected luck” [p. 90]. The lesson that was learned by the CIA and Eisenhower administration, then, was that the CIA was more than equipped to undertake foreign operations that were historically handled by the military. The rise of the CIA vis-à-vis military intelligence after TPAJAX was significant because the operation put “civilian clandestine operational expertise” and “a civilian intelligence agency at the forefront of planning and executing covert operations” [p. 90]. Thus, TPAJAX constituted an important turning point in intelligence history not only because it illustrated the potential for clandestine operations as a tool of conducting U.S. foreign policy, but also because it cemented the CIA’s status as the go-to actor for operating that tool.

Read the documents


(i) Author’s interview, September 12, 2017.