A dead drop is not primarily a hiding place. In the sworn affidavit that led to Robert Hanssen's arrest in 2001, an FBI special agent defined it for the court in a single sentence whose whole weight falls on the last clause: a dead drop is "a prearranged hidden location used for the clandestine exchange of packages, messages, and payments, which avoids the necessity of an intelligence officer and an agent being present at the same time."
The concealment is a means. The absence is the point. And the mechanism that produces that absence is not the cache at all: it is a second, entirely separate location that carries no package and holds nothing worth finding.
Key Takeaways
- The FBI's own definition of a dead drop, filed in federal court in 2001, describes it by what it prevents rather than by what it hides.
- Robert Hanssen's written instructions of October 1985 broke a single handover into five visits by two people, three of them to a signpost that held nothing, so that the two never overlapped at either location.
- Hanssen told his handlers a meeting's real danger was not surveillance but that it would reveal his identity, which he called his "best protection against betrayal" from inside their own service.
- A 1988 letter records what happens without a working signal: he walked to an empty cache, could not tell whether the schedule had slipped or the operation was blown, and wrote back "Empty sites bother me."
- The same convenience that makes a drop site usable makes it an address. Investigators noted that one site had been used at least seven times and sat about a one-mile walk from Hanssen's house.
What does a dead drop actually replace?
That affidavit, filed in the Eastern District of Virginia in February 2001, opens with a glossary, because an affidavit has to be intelligible to the magistrate signing the warrant. The effect is that the tradecraft is on the record in plain terms. Paragraph 14 gives the dead drop, quoted above. Paragraph 15 gives the other half:
A signal site is a prearranged fixed location, usually in a public place, on which an agent or intelligence officer can place a predetermined mark in order to alert the other to operational activity. Such a mark may be made by, for example, chalk or a piece of tape. The operational activity signaled may be the fact that a dead drop has been "loaded" and is ready to be "cleared."
The asymmetry between them is the whole design. A cache can only be interrogated by opening it, which is the single act that would convict whoever performs it. A mark can be read from a moving car by someone with no reason to stop. The signal site is therefore not an accessory: it is the only instrument in the system that can be safely consulted, which means every question either party will ever need to ask has to be phrased so that it fits on a signpost.
Which risk was Hanssen actually avoiding?
The intuitive answer is surveillance: two people converging is a pattern that a watching team can photograph. That is true, and it is not the reason Hanssen gave.
In a letter quoted at paragraph 133 of the affidavit, written to his Russian handlers in late 2000, he refused an offer to meet in person and explained why in terms that have nothing to do with cameras:
Meeting in this country is not really that hard to manage, but I am loath to do so not because it is risky but because it involves revealing my identity. That insulation has been my best protection against betrayal by someone like me working from whatever motivation.
He was an FBI counterintelligence agent, and knew professionally that a foreign service can be penetrated, and that anything its officers know about a source can be sold on by the next Hanssen. A face cannot be recalled once handed over. He had put it more bluntly in a letter of July 1988, quoted at paragraph 91: "I am much safer if you know little about me."
That was not rhetoric. The Justice Department's Office of the Inspector General, reviewing the case in 2003, counted the concealment of his name among the few genuinely competent things he did, describing him as taking "some important steps to maintain his security, such as refusing to reveal his identity to his Russian handlers," while finding much of his other conduct reckless.
The one occasion he broke it shows what the rule was buying. In July 1993 he walked up to a Russian military intelligence officer in the garage of the officer's apartment building, identified himself as an FBI agent, gave the alias he had used for years, and tried to hand over a package. The officer refused it and reported the approach. Russia lodged a diplomatic protest and the FBI opened an investigation. The inspector general called the episode "remarkable for its recklessness." It is the only such approach that review records, and it produced the closest thing to exposure in his entire career.
What did a signal site actually say?
The instructions survive because Hanssen typed them himself and mailed them to a KGB officer's home address in Alexandria in October 1985. The affidavit reproduces them at paragraph 57.
The KGB's package for him was to go under the corner of a wooden footbridge just west of the entrance to Nottoway Park, wrapped in a green or brown plastic trash bag with rubbish packed around a waterproofed inner parcel. Nothing in the instructions concerns hiding the writing, which is a separate discipline with its own history of reagents and heat-developed inks. The signal site was somewhere else entirely: a pictorial pedestrian-crossing signpost on the road outside. Then the protocol:
My signal to you: One vertical mark of white adhesive tape meaning I am ready to receive your package.
Your signal to me: One horizontal mark of white adhesive tape meaning drop filled.
My signal to you: One vertical mark of white adhesive tape meaning I have received your package.
(Remove old tape before leaving signal.)
Laid out by actor, the same instructions look like this:
| Step |
Who moves |
Where |
What is left there |
| 1 |
Hanssen |
Signpost |
Vertical tape: ready to receive |
| 2 |
KGB |
Footbridge cache |
The package |
| 3 |
KGB |
Signpost |
Horizontal tape: drop filled |
| 4 |
Hanssen |
Footbridge cache |
Nothing; the package is taken |
| 5 |
Hanssen |
Signpost |
Vertical tape: package received |
Five visits by two people to two locations, and at no moment are both parties at either place together. Three of the five are to the signpost, which holds nothing and proves nothing. Only two involve the cache, and the second of those happens after the system has already confirmed that the first one worked.
Note what the marks encode. Orientation carries all of the meaning: vertical or horizontal, same tape, same post. That is a smaller vocabulary than the dots and dashes of Morse, which compose freely into any word, and two tape orientations cannot compose at all. The vertical mark also appears twice with different meanings, at steps one and five, and nothing in the symbol distinguishes them. Only position in the sequence does, so both parties have to hold the same state in their heads, and neither can write it down.
Then the housekeeping line in brackets, which is the least glamorous and most important instruction in the document. A stale mark is worse than no mark, because it says the wrong thing with full authority.
What happens when nobody signals?
On 15 July 1988 the KGB received a letter from Hanssen describing a night that had gone wrong. Quoted at paragraph 91, it is the best surviving account of why the signal site exists at all.
He had gone to the cache and found nothing there, and could not tell why. He was working from memory, and his recollection was that the site should have been filled before one in the morning. Either the schedule had slipped, or he had misremembered it, or the operation was blown and he was standing on the evidence. Then this:
I believe Viktor Degtyar was in the church driveway off Rt. 123, but I did not know how he would react to an approach.
The KGB officer he mailed his letters to was, he thought, standing a short distance away, and he could not speak to him. The entire structure of the relationship was built to make that conversation impossible, and it held even when a sentence would have solved the problem.
The rest of the letter asks for exactly the missing component:
Please plan filled signals. Empty sites bother me.
He added that he liked to know before exposing himself, and assumed the other side felt the same. Those two sentences are the argument for the signal site: reading the mark first turns the dangerous half of the operation into something you attempt only once it is already known to be worth attempting. The same letter, still in his own words at paragraph 91, shows the other cost. Because of his job he had to synchronise explanations and flights "while not leaving a pattern of absence or travel that could later be correlated with communication times."
Closing out a dead drop exchange, in the service's own words
A letter Hanssen received on or before 31 July 2000, reproduced at paragraph 132, has the Russian service specifying a full two-way exchange in writing. Every step in it is a signal, and none of them is a meeting.
They would be ready to withdraw his package after eight in the evening on a set date, and only once they had read his signal. They would then load their own package for him, remove his mark and put up their own by nine that same evening. He would then withdraw their package and remove their mark, and, in their words, "that will mean an exchange operation is over." They would check the site again the next evening. If the signal had not been removed by then, they would withdraw their package and re-offer it at a second named site "in each seven days" over the following three weeks.
Nothing there is decoration. Removing a mark is an acknowledgement, a mark left standing is a fault report, and the fallback dates are the recovery procedure. The letter opens by calling the protection of his personal security their "golden rule."
The Ames case shows the same loop failing safely. The Senate Select Committee on Intelligence's 1994 assessment records that on 19 August 1992 the CIA officer Aldrich Ames put a pencil mark on a signal site the pair called HILL in the morning, then left his documents at a cache called GROUND at four that afternoon. The arrangement was that the Russians would erase his mark once they had collected. He went back the next day and the mark was still there, so nothing had been collected. He retrieved his own package. Nobody met, nobody was seen, and a failed exchange was detected and reversed by looking at a pencil line.
The system that hid Hanssen also drew a circle round his house
A drop site has to be somewhere you can plausibly be, and a place you can plausibly be is usually a place you live near. The affidavit turns that into an argument at paragraph 149. Investigators noted that one site had been used at least seven times, and reasoned that such frequency suggested the source lived very close to it or passed it routinely. Hanssen had also told the KGB in 1989 that they could use that particular site at any time. The affidavit then closes the loop: the site sits approximately a one-mile walk from his home address in Vienna, Virginia.
The arrest followed the protocol exactly. A second affidavit, sworn in support of the search warrants, records that at about 4:34 in the afternoon on 18 February 2001 surveillance saw Hanssen at the park sign, placing a piece of white adhesive tape on it, then walking into the wooded park toward a footbridge. He came back out about nine minutes later and was arrested carrying his FBI credentials and a small roll of white adhesive tape. Agents recovered a package wrapped in a taped-up black plastic bag from under the footbridge. The Bureau's press release two days later described him as arrested while "clandestinely placing a package containing highly classified information at a pre-arranged, or 'dead drop,' site."
Signal first, then load, exactly as he had specified in 1985. The design worked. What it could not do was make the man performing it unremarkable, and the inspector general's verdict on more than twenty years of it is deflating: he escaped detection "not because he was extraordinarily clever and crafty" but because of longstanding failures in the FBI's own security programs.
Frequently asked questions
Why does a dead drop need a second location instead of just a hiding place?
Because without a signal site, the only way to learn whether a package is waiting is to open the cache and look, which means repeatedly approaching the single location that would incriminate you. The FBI's 2001 affidavit defines the signal site as a separate prearranged spot, usually public, where a predetermined mark announces that a drop has been loaded and is ready to be cleared. Reading a mark from a passing car is deniable and costs nothing; opening a cache is neither.
What did the marks at a signal site look like?
In the protocol Hanssen sent to a KGB officer in October 1985, every mark was a strip of white adhesive tape on the same signpost, and the meaning was carried entirely by orientation. One vertical mark from him meant he was ready to receive a package; one horizontal mark from the Russians meant the drop was filled; a second vertical mark from him confirmed he had collected it. The instructions ended with a bracketed housekeeping note to remove old tape before leaving a new signal, because a stale mark states the wrong thing with full authority.
Did the two sides ever have to be at the site together?
No, and that is the entire design. The 1985 sequence needed five visits, three of them to the signpost and two to the cache, taken in three alternating turns: his, then theirs, then his again. A schedule the Russian service put in writing in 2000 gave a one-hour evening window in which they would collect his package only after reading his signal, then load their own, then remove his mark and put up theirs. He would collect their package last and take their mark down, and removing that final signal was itself the message that the exchange was complete.
Was avoiding surveillance the main reason for refusing to meet?
Not according to Hanssen. In a letter quoted in the affidavit he said meeting inside the United States was not especially hard to arrange, and that he was reluctant not because it was risky but because it would reveal his identity, which he described as his best protection against betrayal by someone inside the Russian service doing what he was doing. The Justice Department's inspector general later counted his refusal to give his handlers his name among the few effective security measures he actually took.
What happened when a signal was missed?
The clearest surviving case is Hanssen's letter of July 1988, reporting a cache found empty with no way to tell whether the timing had slipped or the operation had been compromised, and noting that he believed his handler was standing in a driveway nearby but could not risk approaching him. He asked for the practice to change, writing that empty sites bothered him. The failure ran in reverse for Aldrich Ames in August 1992: the Russians were to erase his mark once they had collected, and when he found it still standing the next day he recovered his own package rather than leave it sitting.
A tape mark protects the handover but says almost nothing. When the weak point was the message itself rather than the meeting, agent communications went somewhere else entirely: the SOE cryptographer who replaced a memorised poem with a key printed on silk is the other end of that problem.
Sources
- United States v. Robert Philip Hanssen, Affidavit in Support of Criminal Complaint, Arrest Warrant and Search Warrant, U.S. District Court for the Eastern District of Virginia, February 2001. Definitions at paragraphs 14 and 15; the October 1985 drop and signal protocol at paragraph 57; the July 1988 letter at paragraph 91; the July 2000 exchange schedule at paragraph 132; the refusal to meet at paragraph 133; the site-frequency analysis at paragraph 149. Text at https://irp.fas.org/ops/ci/hanssen_affidavit.html - retrieved 10 August 2026
- United States v. Robert Philip Hanssen, Affidavit in Support of Search Warrants, February 2001. Arrest-day surveillance at paragraphs 5 to 8. Text at https://irp.fas.org/ops/ci/hanssen_affidavit2.html - retrieved 10 August 2026
- U.S. Department of Justice, Office of the Inspector General, A Review of the FBI's Performance in Deterring, Detecting, and Investigating the Espionage Activities of Robert Philip Hanssen (unclassified executive summary), 14 August 2003, pages 13 and 15. https://oig.justice.gov/sites/default/files/legacy/special/0308/final.pdf - retrieved 10 August 2026
- U.S. Senate Select Committee on Intelligence, An Assessment of the Aldrich H. Ames Espionage Case and Its Implications for U.S. Intelligence, 1 November 1994, section headed "Ames's Contacts with the KGB". https://irp.fas.org/congress/1994_rpt/ssci_ames.htm - retrieved 10 August 2026
- Federal Bureau of Investigation, "Robert Hanssen", Famous Cases and Criminals, carrying the FBI National Press Office release of 20 February 2001 quoted above. https://www.fbi.gov/history/famous-cases/robert-hanssen - retrieved 10 August 2026
