CIA documents in the HSCA records

[Corrected and revised on 1/17/24; new refs added 1/23/24]

This long-overdue note gives a brief overview of CIA documents that were collected by the House Select Committee on Assassinations (HSCA), a special Congressional committee that investigated the assassinations of John F. Kennedy and Martin Luther King, Jr.

During its 1976-78 investigation, the Committee acquired a massive collection of CIA records, dating from World War 2 all the way up to the 1970s. This HSCA collection is the main source of the CIA records in the John F. Kennedy Assassination Records Collection (JFKARC) at the National Archives today.

The collection, sometimes known as the HSCA CIA Segregated Collection, was first systematically described in a 1992 CIA memo which was written after a month-long survey of the documents by the agency’s History Staff. This post summarizes the survey results, and also notes a few problems in the survey and summary.

I deal here only with HSCA’s CIA records. For a more general overview of JFKARC contents, both from the CIA and other agencies, see here.

Reviewing the collection: The McDonald survey of HSCA records

In January 1992, acting CIA director Robert Gates asked the CIA History Staff to survey the HSCA’s collection of CIA documents and to provide a recommendation on how to handle the documents, already the subject of a massive FOIA court case, and soon to become the subject of sweeping legislative proposals.

CIA’s Chief Historian, J. Kenneth McDonald, responded to Gates in a February memo. After summarizing the content of the records, McDonald recommended that the collection be transferred to the National Archives at its then current level of classification. This recommendation was of course soon rendered moot by the passage of the JFK Act in October 1992.

McDonald’s memo is available in a lengthy CIA record, 104-10337-10006. The MFF copy of this record is here. There are a number of other documents in this records, as well as the memo. A copy of the memo by itself is available here.

The same CIA record also provides a summary description of the HSCA collection, available here. This summary divides the HSCA collection into several major parts: 1) printed records in 63 boxes; 2) microfilm records in box 64; 3) the Lee Harvey Oswald 201 file in 16 boxes; 4) miscellaneous loose folders.

An attachment to McDonald’s memo provides a somewhat more detailed box and microfilm reel list. A copy of this list is here. There are important caveats on McDonald’s memo and survey which researchers should note. See the conclusion of this post for these.

In addition to the HSCA collection of CIA docs, there are two other large sets of CIA docs in the ARC: the Russell Holmes collection and the miscellaneous records assembled by the Assassination Records Review Board (ARRB), the federal board charged with assembling and releasing the documents in the JFKARC. I will have a second post on these doc sets in the near future.

A note on folders and files in the HSCA collection

In addition to the hard copy boxes and microfilm reels covered in the CIA history staff list cited above, documents in the Segregated Collection are also arranged in smaller units inside these big units. Call these smaller units “folders”.

These folders often contain meaningful collections of ordered documents: i.e. “files”. For example, JFK box 40 consists of 39 folders, each of which contains a file on a specific person, compiled by the CIA Office of Security. That means there are 39 files on 39 people, from LUIS ALBERU-SUOTO to TENNENT HARRINGTON BAGLEY (the arrangement is kind-of-alphabetic).

Unfortunately, the McDonald Survey did not provide a folder list, and as far as I can tell, there is no complete folder list for the CIA Segregated Collection available, though there are a number of partial lists in various places.

Early folder/file releases

It would be nice if the documents collected in the segregated collection folders were available as real files, and in the earliest releases by the CIA, before the ARRB was up and running, they were indeed released as such. These early “file” releases are available on the MFF website, though it can take some searching to find them. For example, the file on Alberu-Suoto is here.

Because this was early in the release process, the MFF copies of these files are often heavily redacted. The Alberu-Suoto file mentioned above is a typical example. In addition to plentiful redactions, many documents are omitted from the file, replaced with Document Withdrawal Notices (DWNs).

Now that the documents are almost all released in full, it would be nice to see updated versions of these files. Unfortunately, updated versions are not available.

No later folder/file releases

In fact, there were virtually no folder releases from the Segregated Collection after about 1994, though the 2017-2018 releases from the JFKARC did again produce long, long pdf files that seem to represent whole folders.

Instead, for the duration of the ARRB’s tenure, documents from a wide range of files were released individually. Strangely, it seems this was a deliberate decision by the ARRB. Why did they decide to do this?

Once the ARRB got going in 1994, they chose to require the reprocessing and release of the documents in each folder/file on an individual basis. Each document is identified by an identification aid, called a Reader Information Form (RIF). laid on top of each record. Each RIF includes a record number (RIF number) and a bunch of metadata describing the record (# of pages, date, release state, etc.).

By atomizing files into individual documents, the ARRB was able to require CIA to justify every single document redaction they proposed, and thus exert strong pressure on CIA to release every scrap of information in the doc.

Unfortunately, the cost of this decision was that the orderly folders/files turned into one gigantic pile of numbered documents. In my opinion, these was an extremely high price to pay. Putting the documents back together into files can be done, but it is not a simple task. One must go through the records box by box, folder by folder, and doc by doc.

How does one do this?

Finding records by box or reel and folder

Fortunately, one can reconstitute file folders in the CIA records collection using information from the RIF identification aid.

Look in the “comments” field of each RIF form, and you will see a string of letters and numbers. For example, in the RIF form of document 104-10110-10152, the comments field reads “JFK40 : F1 : 1993.08.02.09:40:37:210060”. (This document is not available online, but you can get the comments data from the MFF JFK Database Explorer.

JFK40 refers to box 40 of the hard copy documents in the HSCA collection of CIA records. F1 means folder 1. The remainder is probably an accession number used in the CIA’s internal database of JFKARC records, indicating the date and time the document was entered in database, plus a number at the end to make sure records with the same date-time stamp are distinct.

Looking back at the Alberu-Souto file, we can see that record 104-10110-10152 is none other than the first page of the file, the log sheet HSCA researchers signed when they drew out a file from the CIA for review. This sheet shows us that Alberu’s file was reviewed by HSCA researcher Ed Lopez on August 15, 1978.

To get a list of the other documents in Alberu’s file, one simply searches through the JFK Database for all the records with the comment string JFK40 : F1. Alas, searches like this require you to download the JFK database excel files from NARA and do it yourself, not a simple task. And there are sometimes other, even more complex problems to solve when figuring out what goes where.

One particularly tricky problem is putting the documents back in the right order. Past the folder level, there is no way of telling the order of documents in the file. In some cases, it seems that the document’s RIF number may mark its order in the file, but there are lots and lots of exceptions. For example, documents withdrawn in the first releases of the files often have completely different RIF numbers than documents that were included in the first releases.

I also suspect that some of these files were simply boxed up in the order preserved in the folder. If there is an early release of the file, one can consult that for document ordering but I don’t think that can always be treated as reliable.

It looks like the HSCA researchers, when reviewing the files, were free to shuffle the file docs around as they wished. The docs weren’t bound or stapled together at all. In some cases CIA paper shufflers shuffled the papers back into reverse chronological order (latest first, earliest last). If this were done consistently, all the files should have the HSCA log sheet first, but in many files the log sheet is randomly placed. In other words, the files were a mess when they were packed up.

By going through the comment strings in the latest releases of all the documents, once can provide “reconstructed” versions of folders/files with all the latest redaction in place. I have done this now for a number of files. Here, for example, is a link to my “reconstruction” of Alberu’s OS file. For another example, see my post on CIA officer J Walter Moore’s OS file here.

Notice that although all the records in these two files have been released in full, not all of these records are available online. Hence I have left DWNs in place where the record is not online. Again, these records are actually available at NARA. I’ve written to NARA asking for copies, and when I get them, I’ll insert them into my “reconstructions” and post them here.

Standardizing references

It is natural to use the abbreviations in the RIF forms for the different parts of the HSCA collection of CIA documents. For example, a box of hard copy folders/docs is simply CIA JFK box XX.

Box 64 was a collection of 72 reels of microfilm. These are identified in the RIF forms as JFK64-1 TO JFK64-72. I will simply call these CIA JFK reel XX. Microfilm reels don’t have folders of course, but there are file markers in the reels, so I will keep using the word “folders” here as well.


Summarizing, this note identifies two sets of CIA documents in the HSCA records: “hard copy” printed documents arranged in 63 boxes (CIA JFK box XX), and 72 reels of microfilm, originally in box 64 (CIA JFK reel XX). Boxes and reels are sub-divided into folders which are often equivalent to one file. Some folders may have multiple files, other folders seem to be more or less miscellaneous collections of documents.

The division between CIA records and HSCA records is not always neat. There are large quantities of CIA correspondence and HSCA researcher notes on CIA documents in the HSCA section of the JFK ARC. These use the HSCA record number prefix “180-“, but in some cases one can actually identify which CIA JFK box these records, now designated HSCA records, were extracted from. In other cases, it would take a lot of research to say which came from where.

It is surely safe to say that the HSCA had access to the vast majority of the CIA records in the Segregated Collection, though there may have been some internal correspondence during 1978-79 that HSCA did not see. When I say “had access”, I don’t mean that every member of the HSCA staff saw every document in the CIA collection. Some documents, such as the IG report on Castro assassination plots, were circulated in multiple editions, including heavily sanitized, lightly sanitized, and completely unsanitized. Who saw what would require careful research.

In addition, the HSCA requested significant quantities of CIA records which they never bothered to look at. Scott Breckinridge, the primary CIA liaison with the HSCA, estimated as much as 20% of the records CIA provided were never checked out.

There are other sources of CIA records in the JFKARC, as I mentioned above, and I will have more to say on these in other notes.

Now for the caveat. Much of this post relies on CIA surveys and research on the HSCA records. It is important to note that a lot of this was done on extremely tight schedules with limited manpower. As weeks stretched into months, years, and decades, CIA (and we, their constant readers) obviously became more familiar with the records. Later familiarity tells us there were errors in earlier studies.

McDonald’s memo on the HSCA records and the February 1992 history staff summary of the collection each has a notable example of such errors. McDonald identifies Clay Shaw as a “highly paid contract source” for the CIA. The staff summary identifies Gilberto Alvarado as someone who “witnessed Cubans passing Oswald cash at a party on the night before the assassination.”

Both of these statements are contradicted by numerous documents in the collection, and confirmed by none. They are certainly errors. This has been pointed out by other researchers, including Paul Hoch, Max Holland, Fred Litwin, and others. The documents say what they say. A summary that varies from what the documents record is in error.

In addition, the history staff box and reel list often conflicts with other such lists, and sometimes with the RIF form information on box and reel content as well. Boxes are renumbered and reordered, and sometimes the entire shebang is completely reboxed. The box and reel list is an interesting overview, but one must go to the individual document data (the comments field specifically) as a whole for a thorough reconstruction.

Other sources on HSCA-CIA docs

Harrelson and Pereira presentation to ARRB
ARRB Final report, Chapter 5