CASE STUDY

The Cleveland Torso Murders: The True Story of the Mad Butcher of Kingsbury Run

By Domonique Luchin · Published November 4, 2026 · 16 min read

Watch: At Least 12 Dead. No One Ever Tried. Cleveland's Torso Murders #shorts

On September 23, 1935, two boys playing near the base of a steep slope called Jackass Hill, where East 49th Street dead-ends at Kingsbury Run on Cleveland's east side, found two bodies in the weeds. Both men had been decapitated. Their heads were buried nearby. One of them was wearing nothing but socks. Neither had been killed where he lay.

Over the next three years, Cleveland would find at least ten more. Men and women, almost all of them poor, most of them never identified, every one of them beheaded. The city's newspapers called the unknown killer the Mad Butcher of Kingsbury Run. The man who led the city's police and fire departments during the hunt was Eliot Ness, fresh from his fame chasing Al Capone's operation in Chicago. And at the end of it, after some 9,100 people were questioned, a jailhouse confession, a secret polygraph and a city neighborhood burned to the ground, no one was ever tried for a single one of the killings.

This is the true story of the Cleveland Torso Murders, told from the record: the victims, the places they were found, the investigators, the two men police suspected most, and the DNA work that is still trying, nearly a century later, to give the victims back their names.

On The Sealed Docket, host Domonique Luchin works from the record: real names, real documents, real footage. This write-up draws on the Encyclopedia of Cleveland History published by Case Western Reserve University, the Cleveland Police Museum's history of the case, reporting from WOIO (Cleveland 19 News), the DNA Doe Project's own case files, and the public record summarized on Wikipedia, which in turn relies heavily on historian James Jessen Badal's book In the Wake of the Butcher. A note before you read on: this is a case about murder and dismemberment. The details here are kept to what the record supports and what you need to understand the case.

Kingsbury Run in the Great Depression

A ravine on the edge of the city

Kingsbury Run was a creek bed and ravine that ran from around East 90th Street and Kinsman Road down to the Cuyahoga River, according to the Encyclopedia of Cleveland History. By the 1930s it was crisscrossed by railroad tracks, and its slopes and flats held shantytowns, the clusters of shacks that Depression-era Americans called "Hoovervilles." Men who rode the freight trains in and out of Cleveland camped there. So did families who had nowhere else to go.

Just east of the Run was a district known as the Roaring Third, a stretch of bars, gambling dens, brothels and cheap rooming houses. Most of the Torso Murderer's victims came from that world. That matters for understanding why the case was so hard. The victims were people whose absence few people reported. Many were transients with no address, no family nearby and no fingerprints on file.

A killer who removed every clue

The Encyclopedia of Cleveland History describes body parts of what was thought to be seven men and five women, nearly all found in or near Kingsbury Run, between September 1935 and August 1938. All were decapitated. Most were also dismembered. Investigators noticed how precise the cuts were, and the police came to believe the killings were the work of one person with real knowledge of anatomy.

In many cases, according to the coroner's findings summarized on Wikipedia, the decapitation itself was the cause of death. Some bodies had been treated with a chemical that turned the skin red, tough and leathery. Some were not found until months, or in one case about a year, after death. Heads were often never recovered. In an era before DNA, with forensic science still young, removing the head meant removing the face, the teeth and often the only realistic way to put a name to a victim.

The Victims of the Mad Butcher

The "official" count used by most researchers is twelve victims, though some investigators believed there were more. Only two were positively identified at the time, and a third tentatively. Here is what the record says about each.

Edward Andrassy and John Doe I (September 23, 1935)

The first two bodies were the pair found at the foot of Jackass Hill. One was identified by his fingerprints as Edward Andrassy, 29, who had once worked as an orderly in the psychiatric ward at Cleveland City Hospital but was unemployed when he died. The coroner found rope burns on his wrists and ruled that he died of decapitation. He had been dead two or three days. The second man, found nearby, was never identified. His body appeared to have been soaked in oil and set on fire after death. He became known as John Doe I.

Florence Polillo (January 26, 1936)

Florence Genevieve Polillo, 44, was found dismembered, wrapped in paper and packed into half-bushel baskets behind a building on East 20th Street. Her head was never found. She was identified by her fingerprints. She would become central to the case years later, because the one man ever charged with a Torso killing was charged with hers.

The Tattooed Man (June 5, 1936)

In June 1936, the headless body of a young man was found near railroad tracks by an old freight shed. His head turned up about 1,500 feet away. There was no blood on the ground, which told investigators he had been killed elsewhere. He was estimated to be in his early twenties, and he had six tattoos, including the names "Helen and Paul," a heart and anchor, a butterfly, the initials "W.C.G." and the comic-strip character Jiggs. That summer Cleveland was hosting the Great Lakes Exposition, and the police displayed his death mask there. According to Wikipedia's summary of Badal's research, thousands of people viewed it. No one recognized him. He is still known only as the Tattooed Man.

John Doe III (July 22, 1936)

The badly decomposed, headless body of a man was found near a homeless camp in the Big Creek area of Brooklyn, west of Cleveland. He was the only known victim on the city's West Side, and the only one investigators believed had been killed where he was found. His remains were too decayed for fingerprints.

John Doe IV (September 10, 1936)

A man waiting for a freight train spotted the two halves of a man's torso floating in a stagnant pool in Kingsbury Run near East 37th Street. Police found a felt hat from a shop in Bellevue, Ohio, with what appeared to be blood on it, and a bloodied work shirt wrapped in newspaper. A fire crew dredged the creek. The head was never found and the man was never identified.

Jane Doe I (February 23, 1937)

The upper half of a woman's torso washed up on the shore of Lake Erie near Euclid Beach. Three months later the lower half came ashore at East 30th Street. Her head, arms and legs were never found. She was found almost exactly where another woman's remains had washed up in 1934, a point we come back to below.

Jane Doe II, possibly Rose Wallace (June 6, 1937)

Beneath the Lorain-Carnegie Bridge, searchers found the partial skeleton of a woman in a rotting burlap bag along with a newspaper from June 1936. She had been dead about a year. Police tentatively identified her as Rose Wallace, 40, who had disappeared in August 1936 from the same bar where Florence Polillo had been a regular. Dental work was considered a close match and Wallace's son said he was certain it was his mother. But the dentist who did the work had died, and the timing did not quite fit, so officially she remains unidentified.

John Doe V (July 6, 1937)

The upper part of a man's torso, wrapped in a burlap chicken-feed sack, was found floating in the Cuyahoga River in the Flats just below Kingsbury Run, along with his thighs. His head and heart were never found.

Jane Doe III (April 8, 1938)

A woman's leg was found in the Cuyahoga River. A month later, two burlap bags holding the rest of her torso were pulled from the river near the West 3rd Street Bridge. Her head and arms were never found. She was the only victim in whom the coroner found morphine, a detail that would later feed suspicion of a doctor.

Jane Doe IV and John Doe VI (August 16, 1938)

The last two "official" victims were found on the same day. Men searching for scrap metal at a dump at the end of East Ninth Street found a woman's dismembered body wrapped in rags, brown paper and cardboard. For once, her head and hands were with the body. Nearby on the lakefront, the remains of a man were found in plain view of Cleveland City Hall, where Eliot Ness had his office. Many writers have read that placement as a taunt. Neither victim has ever been identified.

The Possible Victims: The Lady of the Lake and Beyond

The official twelve may not be the full count. Lead detective Peter Merylo believed there were more victims in the Cleveland, Youngstown and Pittsburgh areas.

The strongest candidate is the woman known as the Lady of the Lake. On September 5, 1934, a year before the first official victims, the lower half of a woman's torso washed up on the Lake Erie shore just east of Bratenahl. She was found virtually in the same spot as Jane Doe I in 1937, and both women's remains had a chemical on the skin believed to be a form of lime. Some researchers think the killer meant to speed up decomposition and used the wrong kind of lime, which preserved the bodies instead.

Another is Robert Robertson, 41, found decapitated on Davenport Avenue on July 22, 1950, after being dead six to eight weeks. Newspapers tied him to the Torso Murderer, but detectives treated it as an isolated crime.

Investigators also looked hard at Pennsylvania. Between 1921 and 1942, nine people, eight of them unidentified, were found dead and dismembered in swamps and train yards near Pittsburgh, a city linked directly to Cleveland by a Baltimore and Ohio Railroad line. A headless man was found in a boxcar in New Castle, Pennsylvania, on July 1, 1936, and three more headless victims turned up in boxcars near McKees Rocks on May 3, 1940. Merylo was convinced the cases were linked. None of those links has ever been proven.

Eliot Ness Takes the Case

The Untouchable in Cleveland

Eliot Ness came to Cleveland after his years as a federal Prohibition agent in Chicago, where his squad of "Untouchables" had gone after Al Capone's breweries. In December 1935, a few months after the first two bodies were found, he was named Cleveland's Safety Director, the official over the police and fire departments. He spent his first years cleaning up police corruption and organized gambling. The Torso Murders became the case his reputation could not fix.

The day-to-day hunt belonged to detectives like Merylo, who worked the case for six years and, according to the Encyclopedia of Cleveland History, described the killer as coldly efficient and as relentless as an executioner. Merylo even went undercover as a hobo, riding the rails on the theory that the killer was a transient who moved with the freight trains. Cuyahoga County Coroner Samuel Gerber examined the remains and studied the cuts. According to Wikipedia, authorities questioned around 9,100 people over the course of the investigation.

Burning the shantytowns

Two days after the last two official victims were found in August 1938, the investigation took its most controversial turn. Ness personally led raids on the shantytowns of Kingsbury Run and the Flats. Residents were rounded up and fingerprinted, and the shacks were burned to the ground. Ness's reasoning, according to the Cleveland Police Museum, was twofold: the fingerprints would make any future victim easier to identify, and clearing the camps would remove the killer's pool of victims.

Four days after the burning, on August 22, 1938, Ness sent six two-man teams across a wide area of the city, from the Cuyahoga River to East 55th Street and Prospect Avenue, posing as fire inspectors. They found no new evidence about the killer. What they did find, according to Badal's account, were hundreds of families living in dangerous fire traps without toilets or running water. The raids drew sharp public criticism, but the Encyclopedia of Cleveland History notes that the official series of killings ended after them.

The Suspects

Of the thousands questioned, two men stood out. Neither was ever convicted of any Torso killing.

Frank Dolezal: a confession and a jail cell

Frank Dolezal, a Cleveland man born in 1887, had lived with Florence Polillo and was said to have known Edward Andrassy and Rose Wallace. His odd behavior and his collection of knives drew attention. He was arrested in the summer of 1939, not by Ness's city police but by Cuyahoga County Sheriff Martin O'Donnell's office. The Associated Press, in a report carried by The New York Times on July 8, 1939, said the prisoner had admitted one torso slaying.

Dolezal told investigators he had killed Polillo in self-defense. Then he took the confession back and said it had been beaten out of him. On August 24, 1939, before any trial, he was found hanged in his cell in the county jail. His death was officially ruled a suicide, and rumors that he was killed in custody have never gone away. He is now generally considered to have had nothing to do with the Torso Murders, and Wikipedia's summary says he was posthumously exonerated. Per the Encyclopedia of Cleveland History, there was never enough evidence to bring a case against anyone else.

Dr. Francis Sweeney: the man Ness believed did it

The suspect Ness himself focused on was Dr. Francis Edward Sweeney, born in 1894. Sweeney was a World War I veteran who had served in a medical unit that performed amputations in the field. After the war he drank heavily, and by 1934 his drinking had ended his marriage. In 1929 he had been a surgical resident at St. Alexis Hospital, in the Kingsbury Run area. His medical training fit the investigators' belief that the killer knew anatomy.

In 1938, Ness had Sweeney held and questioned. According to Cleveland Magazine's reporting, Sweeney was so drunk that he was kept in a hotel room for three days until he was sober enough to be interviewed. The polygraph pioneer Leonarde Keeler tested him, and according to Wikipedia's sources Sweeney twice "failed to pass," and Keeler told Ness he believed Sweeney was the man.

But a polygraph is not evidence a jury can convict on, and the rest of the case against Sweeney was circumstantial. There was also a political problem: Sweeney was a cousin of U.S. Congressman Martin L. Sweeney, one of Ness's loudest critics, who had publicly attacked him for failing to catch the killer. Sweeney was never charged. He later committed himself to an institution, and from there he sent taunting postcards to Ness and his family into the 1950s. The Encyclopedia of Cleveland History holds one sent to Ness in 1954 by an asylum inmate some believe was the Torso Murderer. According to Badal, the postcards stopped only after Ness died in 1957. Sweeney died at a veterans hospital in Dayton, Ohio, on July 9, 1964.

Willie Johnson and the 1942 trunk murder

Coroner Gerber also pointed to Willie Johnson, who in June 1942 was seen disposing of a trunk that held the torso of 19-year-old Margaret Frances Wilson. A bootleg-joint operator later identified Johnson in a line-up as a man he had seen with Rose Wallace in August 1936. Johnson was convicted of Wilson's murder and executed on March 10, 1944, but he was never conclusively tied to the Torso killings.

One killer, or several?

Not everyone accepts that the official victims were the work of one person. A 1997 theory, discussed by the Cleveland crime writer John Stark Bellamy II, held that there may have been no single Torso Murderer at all, on the assumption that the autopsy results were inconclusive. The prevailing view among police at the time, and among most historians since, remains that one person did it.

The Black Dahlia Question

In December 1938, someone claiming to be the Torso Murderer wrote to Ness, saying he had moved to California, killed a woman there and buried her head in Los Angeles. The letter writer called himself "DC," as in Doctor of Chiropractic. Los Angeles investigators dug where the letter pointed and found animal bones.

Nine years later, the murder of Elizabeth Short, the Black Dahlia, brought the question back. Short's body had been cut in half and cleaned, and some writers have argued for a link to Cleveland. But Short was not decapitated, which was the one constant in the Cleveland killings, and her murder came nearly a decade after the letter. The original Los Angeles investigators looked at the Cleveland case and discounted a connection. Outside of speculation, nothing ties the two cases together.

Why the Cleveland Torso Murders Were Never Solved

The record points to a handful of reasons that compounded each other:

The victims had no names. Of the twelve official victims, only Edward Andrassy and Florence Polillo were positively identified, and Rose Wallace only tentatively. Without knowing who a victim was, detectives could not trace who they had been with.
The heads were missing. Decapitation removed faces and teeth, and many heads were never recovered at all.
Time destroyed the evidence. Some remains sat in water, dumps or weeds for months before anyone found them.
Forensics were limited. In the 1930s there was no DNA testing, and fingerprints only helped when a victim had been printed before.
The main suspects were never tried. Dolezal died in jail before trial. Sweeney was never charged, and the only evidence against him that Ness found convincing was a polygraph.
The victims' world was mistrusted and transient. People in the camps and the Roaring Third moved constantly, and many had good reason not to talk to police.

The DNA Work Happening Now

The case is not frozen. In August 2024, the Cuyahoga County Medical Examiner's Office began exhuming unidentified Torso victims from a potter's field in partnership with the nonprofit DNA Doe Project, which uses investigative genetic genealogy to identify the dead. WOIO reported on August 21, 2024, that the first two cases sent for testing were the Tattooed Man, listed by the project as Kingsbury Run John Doe #4, and a second unidentified man listed as Kingsbury Run John Doe #6. The project's numbering is its own and does not match the order used in older accounts; its case page places John Doe #6 west of East 37th Street in Kingsbury Run on September 10, 1936. WOIO reported that finding where the victims were buried in the potter's field had been a challenge.

As of the DNA Doe Project's case page, last updated October 30, 2026, the John Doe #6 case is still listed as pending, and no identification has been announced. Even if the killer is never named, the work could one day give the Tattooed Man the name thousands of Clevelanders could not give him in 1936.

Where to See the Case Today

The Cleveland Police Museum keeps an exhibit on the Torso Murders, including death masks of victims including Edward Andrassy, Florence Polillo, the Tattooed Man and Jane Doe II. Cleveland State University's Cleveland Memory Project holds the Cleveland Press photo collection, which includes images of the shantytown fires and the postcards sent to Ness. Kingsbury Run itself is still there, crossed today by rapid-transit and rail lines.

Cleveland Torso Murders Timeline

September 5, 1934: The "Lady of the Lake" washes ashore near Bratenahl (possible victim).
September 23, 1935: Edward Andrassy and John Doe I are found at Jackass Hill, Kingsbury Run.
December 1935: Eliot Ness becomes Cleveland's Safety Director.
January 26, 1936: Florence Polillo's remains are found on East 20th Street.
June 5, 1936: The Tattooed Man is found; his death mask is later shown at the Great Lakes Exposition.
July 22, 1936: John Doe III is found in Brooklyn, west of Cleveland.
September 10, 1936: John Doe IV is found in a pool near East 37th Street.
February 23, 1937: Jane Doe I washes ashore at Euclid Beach.
June 6, 1937: Jane Doe II, possibly Rose Wallace, is found under the Lorain-Carnegie Bridge.
July 6, 1937: John Doe V is found in the Cuyahoga River.
April 8, 1938: Jane Doe III's remains begin turning up in the river.
August 16, 1938: Jane Doe IV and John Doe VI are found, the latter in view of City Hall.
August 18, 1938: Ness leads raids that burn the Kingsbury Run shantytowns.
1938: Dr. Francis Sweeney is questioned and polygraphed; he is never charged.
December 1938: A letter claiming to be from the killer points police to Los Angeles.
August 24, 1939: Frank Dolezal is found hanged in the county jail before trial.
1957: Eliot Ness dies; the postcards stop.
July 9, 1964: Francis Sweeney dies in Dayton, Ohio.
August 2024: Exhumations begin for DNA identification of unidentified victims.
Today: No one has ever been tried for any of the killings.

Watch the Cleveland Torso Murders Case File on The Sealed Docket

Domonique Luchin breaks down the twelve victims, Eliot Ness and the trial that never happened in "At Least 12 Dead. No One Ever Tried. Cleveland's Torso Murders". For the Los Angeles case investigators once tried to connect, read the true story of the Black Dahlia. For another Depression-era hunt with a famous lawman at its center, read the true story of John Dillinger and the Biograph Theater; for another case with no ending, read what happened to Judge Crater. Then browse the full Sealed Docket case index, every episode, and more about the host.

Follow for more case files. New real-case files drop every week on the @TheSealedDocket YouTube channel. Subscribe so the next one finds you.

Related Case Studies from The Sealed Docket

Frequently Asked Questions

What were the Cleveland Torso Murders?

A series of at least twelve killings in Cleveland, Ohio, between September 1935 and August 1938. Every victim was decapitated and most were dismembered, and nearly all were found in or near Kingsbury Run, a ravine on the city's east side. The killer was called the Mad Butcher of Kingsbury Run.

Were the Cleveland Torso Murders ever solved?

No. The killer was never identified, and no one was ever tried for any of the twelve official killings.

Who were the main suspects in the Cleveland Torso Murders?

The two main suspects were Frank Dolezal, who confessed to killing Florence Polillo, retracted the confession and died in the county jail before trial in August 1939, and Dr. Francis Sweeney, a surgeon and World War I medical veteran whom Eliot Ness believed was the killer. Sweeney was never charged.

What did Eliot Ness do in the Torso Murders case?

As Cleveland's Safety Director, Ness oversaw the investigation. In August 1938 he led raids that burned the Kingsbury Run shantytowns and fingerprinted their residents, and he had Dr. Francis Sweeney secretly questioned and polygraphed.

How many victims did the Cleveland Torso Murderer have?

Twelve victims are officially attributed to the killer, thought to be seven men and five women. Some investigators, including detective Peter Merylo, believed there were more, including the Lady of the Lake found in 1934.

Were the Cleveland Torso Murders connected to the Black Dahlia?

A connection has been suggested but never proven. Elizabeth Short was not decapitated, and Los Angeles investigators discounted a link.

Have any Torso victims been identified with DNA?

Not yet. In August 2024 the Cuyahoga County Medical Examiner and the DNA Doe Project began exhuming unidentified victims, including the Tattooed Man, for genetic genealogy testing. As of late October 2026 the cases are listed as pending.

Sources

← All articles · Case index · All episodes