Skip to main content

Sometime in 1961, in a windowless room at CIA headquarters in Langley, Virginia, a veterinary surgeon picked up a scalpel and made an incision along the belly of a gray-and-white cat. The surgery took about an hour. When the surgeon was done, the cat had a small microphone tucked inside its ear canal, a radio transmitter sewn into the base of its skull, and a thin wire antenna running from the back of its neck all the way down its spine, hidden in its fur. They sewed it back up, gave it some time to recover, and started training it.

The cat’s job, when the training was done, was to walk over to two Soviet agents sitting on a park bench and listen to their conversation.

Now sit down, because this one is going to take a minute.

The program had a name. They called it Acoustic Kitty. It was run by the CIA’s Directorate of Science and Technology, the same office that had previously produced robotic pigeons, miniature listening devices the size of a tooth filling, and at least one umbrella designed to deliver poison. According to former CIA officer Victor Marchetti, who was in a position to know, the program ran for about six years and cost American taxpayers roughly twenty million dollars. That is somewhere north of $190 million in 2026 money. Spent on a cat.

The technical problem, as the engineers framed it, was relatively straightforward. The Soviet embassy was hard to bug. The KGB swept the rooms. They swept the cars. They swept the trees outside the windows. What they did not sweep — because nobody on Earth would think to sweep one — was a stray cat. A cat could walk into a courtyard, sit under a bench, lick its paw, and broadcast forty minutes of classified conversation back to a van parked half a block away. The engineering, in theory, was solved. They had the microphone. They had the transmitter. They had the battery. They could shrink it all down small enough to fit inside an animal. The only remaining variable, as one CIA officer put it years later, was the cat.

The cat, it turned out, was the entire problem.

When the engineers tested their first surgically modified cat in the lab, it worked beautifully. The microphone picked up sound. The transmitter sent a clean signal. The cat seemed unbothered. They moved the testing outside. That is where things began to fall apart. The cat, being a cat, did what cats do. It got distracted. It wandered. It got hungry and walked off the job to look for food. According to Marchetti’s account — and his exact words, which I am not going to clean up because they are too good to clean up — “They found he would walk off the job when he got hungry, so they put another wire in to override that.” Read that sentence again. The Central Intelligence Agency conducted a second surgery on a cat specifically to install a wire that would suppress the cat’s hunger so that the cat would stop being a cat about it.

After five years of training, surgery, retraining, and at least one additional procedure to teach the cat not to think about lunch, the program was ready for its first live operational test. The target was a pair of men sitting in a park outside a Soviet compound on Wisconsin Avenue in Washington, D.C. The handlers parked a surveillance van across the street. They opened the door. They let the cat out. The cat walked toward the curb. A taxi turned onto Wisconsin Avenue. The cat did not see the taxi. The taxi did not see the cat. The first and only field deployment of Acoustic Kitty was over before the animal reached the other side of the road.

A CIA operative reportedly got out of the van, walked into the street, and collected what was left of the cat — partly out of decency, mostly to make sure the Soviets did not get their hands on twenty million dollars worth of miniaturized microphone.

A few months later, in March of 1967, a CIA memorandum closed the program out. The document was titled “Views on Trained Cats” — and then two pieces of the title are still redacted, sixty years on. The memo, which was finally declassified in 2001 and lives today at the National Security Archive at George Washington University, did not exactly admit defeat. It congratulated the team for proving “that cats can indeed be trained to move short distances.” It called this “in itself a remarkable scientific achievement.” It then concluded, with the kind of bureaucratic understatement that is its own art form, that “the environmental and security factors in using this technique in a real foreign situation” made it “impractical.” Translation: the cat got hit by a cab and we are quietly walking away from this one.

Here is where the story gets weirder, because it does not actually end with the memo. In 2013, a former director of the CIA’s Office of Technical Service named Robert Wallace gave an interview in which he claimed the taxi story was wrong. According to Wallace, the cat was never killed. The cat was retrieved, the equipment was surgically removed a second time, and the animal “lived a long and happy life afterwards.” Wallace did not produce the cat. He did not produce a date or a name or a record. He just said it. Jeffrey Richelson at the National Security Archive, who has spent more time inside CIA files than most of the people who wrote them, was politely skeptical. “I’m not sure for how long after the operation the cat would have survived,” he said, “even if it hadn’t been run over.”

So here is what we know, and here is what we do not know. We know the program existed. We know it was run out of the Directorate of Science and Technology. We know there were surgeries — at least two of them, on the same animal. We know there was a field test on Wisconsin Avenue. We know that in March of 1967, a senior officer at the Central Intelligence Agency wrote a memo with the words “Views on Trained Cats” at the top of it and meant every word of it sincerely. The document is online. You can read it. About half of it is still blacked out.

What we do not know is the cat’s name. Or whether there were others. The memo refers to “cats” in the plural. The records do not say how many of them there were, or what happened to the rest of them, or where the second surgeries were performed, or who paid the vet bills. The cost figure of twenty million dollars comes from Marchetti, who left the agency in 1969 and wrote one of the most aggressively litigated books in CIA history. The CIA has never confirmed the figure. They have also never denied it.

The document is still there. About four pages, if you don’t count the redactions. The title is “Views on Trained Cats.” It’s free to read online.

I’m just telling you what’s in the file.


Conspiracy Confirmed

Leave a Reply