Content note: this piece describes the disappearance and murder of several young children between 1979 and 1996, and — at its center — the wrongful imprisonment of an innocent man for seventeen and a half years based on DNA evidence that was later proven false. It also discusses a contested, unproven suspect theory raised by a journalist’s investigation; that theory involves disputed DNA results and an individual who has never been charged with any crime. Where the evidence is uncertain, this piece says so plainly, and does not present suspicion as guilt.
The Man They Let Out
Look. I want to start this one somewhere unusual — not with a crime scene, but with a release.
On June 4, 2009, a man named Sugaya Toshikazu walked out of custody in Japan after seventeen and a half years. He had gone in as a thirty-nine-year-old daycare bus driver. He came out a fifty-six-year-old man who had spent nearly two decades of his life in prison for the murder of a four-year-old girl he did not kill. The DNA evidence that convicted him — evidence a court had treated as close to unimpeachable in 1993 — was retested with better technology in 2009 and did not match him. Not partially. Not ambiguously. It did not match him at all.
I need you to sit with the shape of that sentence for a second, because it is the spine of this whole piece: a court sentenced a man to life in prison, substantially on the strength of a DNA test, and the DNA test was wrong. Not fabricated by a villain twirling a mustache — wrong the way flawed science is wrong, applied with total confidence by people who believed in it completely. Sugaya Toshikazu is one of the clearest, most devastating cautionary tales in the modern history of forensic evidence. And his story is only the half of what happened in a narrow stretch of land on the Tochigi–Gunma border between 1979 and 1996, because while he sat in prison for one murdered child, at least three others — maybe four — were taken the same way, in the same twenty-kilometer radius, and nobody has ever been convicted for any of them.
Five Girls, One Twenty-Kilometer Stretch of Land
Here is what happened, in order, across seventeen years, all within a narrow radius straddling the border of Tochigi and Gunma prefectures, north of Tokyo.
August 3, 1979 — Maya Fukushima, 5. She was playing, as she did most days, in the grounds of Yagumo Shrine, right next to her own house in Ashikaga. Her father came home from work for lunch and saw her there, playing, and spoke to her — that is the last confirmed sighting anyone has ever given investigators. When her mother went out to look for her a little after two that afternoon, she was already gone. She was found six days later, on August 9, near the Watarase River, her body left inside a rucksack. The physical evidence and witness testimony were both thin enough that the investigation stalled almost immediately, and the case eventually passed its statute of limitations with no one ever identified.
November 17, 1984 — Yumi Hasebe, 5. Her father, Hasebe Hideo, then forty, had taken the whole family — his wife, five-year-old Yumi, and their four-year-old son — to a pachinko parlor called Daiuchū in Ashikaga’s Yamakawa neighborhood. Pachinko parlors are loud, bright, adult-absorbing places, built around rows of machines that hold a grown-up’s full attention for long stretches at a time, and it was a little after five in the evening before Hideo looked up and realized his daughter was no longer where he’d last seen her. She was gone. Her skeletal remains were found on March 8, 1986 — almost a year and a half later — roughly a mile and a half from the family’s home.
September 15, 1987 — an eight-year-old girl, referred to in most reporting as Tomoko Oosawa. She left her house in what was then Ojima Town, Niitta District, Gunma — now absorbed into Ota City — and never came home. Investigators focused for a time on a man seen on a bicycle with her shortly before she disappeared, but the lead went nowhere; there simply wasn’t enough physical evidence or witness detail to build a case around it. Her skeletal remains were found on November 27, 1988, in the riverbed of the Tone River. Her case reached its statute-of-limitations deadline on September 15, 2002 — fifteen years to the day.
May 12, 1990 — Mami Matsuda, 4. She disappeared from outside a pachinko parlor in Ashikaga. Her body was found the very next day, in the grass along the bank of the Watarase River. Male bodily fluid was recovered from her underwear. This is the case that would consume the next twenty years of Japanese criminal justice history, because this is the case police decided they had solved.
July 7, 1996 — Yukari Yokoyama, 4. She was inside a pachinko parlor in Ota City, Gunma, with her family — playing on her own nearby, the way small children do while adults feed coins into machines a few feet away. The parlor’s own security cameras caught what happened next: a man approaches her, not behaving like someone there to gamble, and speaks to her in a way witnesses and investigators later described as unsettlingly familiar — as though he already knew her, or wanted her to think he did. Witnesses who saw him described a man roughly 158 centimetres tall, wearing loose knickerbocker-style trousers, sandal-like footwear, and sunglasses. Separately, someone reported seeing a girl who looked about four years old being put into a white passenger car around the time Yukari vanished. She has never been found. Hers is the only one of the five that remains, technically, a missing-person case rather than a confirmed murder, because there is no body — which also means it is the only one of the five that is not yet legally time-barred. There is a reward attached to it that has grown into the hundreds of millions of yen over the years. As of 2026, nearly thirty years on, it has never been claimed, and her family is still searching.
Five girls, aged four to eight. One narrow strip of land. Seventeen years. Japan’s own National Police Agency has, at various points, acknowledged the obvious possibility that one person did this. It has never said so as a formal, public conclusion. And for a long time, one of these five cases was treated as closed — which is, in its own way, the most disturbing detail in the entire file.
“Your DNA Matched”
I want to walk you through exactly how Sugaya Toshikazu ended up confessing to a murder he did not commit, because it is not a story about a cruel confession extracted through violence. It is worse than that, in a specific way: it is a story about how easily an innocent person can be talked into believing the evidence against them, when the people holding that evidence are absolutely certain it’s real.
Sugaya was taken into police custody on December 1, 1991 — a year and a half after Mami Matsuda’s murder — in connection with the male bodily fluid recovered from her underwear. Investigators told him his DNA type had matched the sample. He denied it. And then, within hours of being told that a scientific test had already proven his guilt, he stopped denying it and confessed. He was formally arrested the next day and indicted on December 21, 1991.
The DNA test in question used a method called MCT118 typing — an early forensic DNA technique, state of the art for 1991, but one that relied on reading electrophoresis images that were, by later admission, unclear, and on procedures that were far less rigorous than what would become standard practice within the following decade. Both the sample from Matsuda’s underwear and Sugaya’s own sample were read as a “16-26” type. On the strength of that match and his confession, the Utsunomiya District Court sentenced him to life imprisonment on July 7, 1993. Higher courts upheld the conviction in 1996 and again in 2000, closing off his ordinary avenues of appeal.
He would spend the next seventeen and a half years in prison for it.
The Reporter Who Would Not Let It Go
His name is Shimizu Kiyoshi, and I want to give him real space in this piece, because without him, Sugaya Toshikazu very plausibly dies in prison for a crime he did not commit.
Shimizu was a television reporter who began digging into the Ashikaga conviction around 2007 — roughly sixteen years after Sugaya’s arrest, when most of the country had long since filed the case away as solved. He started asking the kind of questions that make institutions uncomfortable: why did a confession obtained within hours of being told about a DNA match hold up so well, when false confessions obtained the same way have collapsed elsewhere? Why was the original DNA testing method — MCT118, already considered outdated and imprecise by the 2000s — never revisited with newer technology? By 2008 he was reporting publicly on inconsistencies in the confession and problems with the original DNA analysis.
Sugaya’s defense team had been pushing for a retest since 2002. The Utsunomiya District Court rejected that request in February 2008. But the Tokyo High Court did something unusual: it agreed to allow a new DNA analysis, conducted independently by two forensic scientists — one nominated by the prosecution, one by the defense, a structure specifically designed so neither side could claim the result was rigged in the other’s favor. Using modern STR-based testing, both scientists reached the same conclusion: the DNA from the crime scene did not match Sugaya Toshikazu. Not at a few loci. At multiple loci, and confirmed independently through both autosomal and Y-chromosome analysis.
The Tokyo High Public Prosecutors Office suspended his sentence and released him on June 4, 2009. A formal retrial opened that October in Utsunomiya — and in an extraordinary reversal, the prosecution itself stood up in court and argued for his acquittal. On March 26, 2010, the court delivered its verdict, and I want you to read the actual language the judges used, because it is about as unambiguous as a Japanese court gets: it is obvious to anyone that Sugaya is not the perpetrator of this crime. The court further found that his original confession had been false testimony, produced through the pressure of interrogation, not through guilt.
Seventeen and a half years. Gone. Because a 1991-era test read a blurry image with more confidence than the image deserved, and because a frightened man, told with total institutional certainty that science had already convicted him, believed it too.
If Not Him, Then Who — and This Is Where I Need to Be Careful With You
Here is where this case stops being a clean story, and I am not going to pretend otherwise.
Shimizu didn’t stop at proving Sugaya’s innocence. His reporting, later expanded into an award-winning book called The Murderer Is There — which won Japan’s Shincho Documentary Prize and the Mystery Writers of Japan Award, and is regarded as a landmark of Japanese investigative journalism — argues that all five of these cases, not just Matsuda’s, were the work of a single perpetrator who was never Sugaya at all, and who has never been arrested.
Multiple eyewitnesses across these cases described a man with a distinctive walk and appearance — wearing knickerbockers, a hat, sunglasses — who reportedly resembled the cartoon character Lupin III closely enough that he became known in the reporting and in public discussion of the case by the nickname “Lupin.” This isn’t a detail that appears in just one case file: the man on the Ashikaga security camera at Yumi Hasebe’s pachinko parlor and the man Yukari Yokoyama’s family’s pachinko parlor caught on its own security footage in 1996 — twelve years apart — were both described in almost identical terms: knickerbocker-style trousers, sandal-like footwear, sunglasses, a height around a hundred and fifty-eight centimetres. That consistency, across more than a decade and multiple prefectures, is the single strongest thread tying these cases together as the work of one person, and it’s also exactly the kind of pattern that made Shimizu believe he was looking at a series rather than five coincidences. Shimizu’s investigation identified a specific individual he believed matched these eyewitness descriptions, and arranged for a DNA sample connected to that individual to be tested by Honda Katsuhiro, a University of Tsukuba forensic professor — the same defense-nominated scientist involved in Sugaya’s exoneration.
And here is the part I most need you to hold onto: the DNA results were not clean. By some accounts, an initial test reportedly showed a full match. A separate examiner’s analysis reportedly found the samples did not match at all. The results, as reported, were inconsistent — not a confirmed match, not a confirmed exclusion. This individual has never been named publicly in a way independent outlets have confirmed, has never been arrested, has never been formally charged, and has never had a chance to answer any of this in a court that could actually test the evidence against the standards a real trial demands.
I am telling you this theory exists because it is a real, significant, publicly documented part of this case’s journalism — you would find it within minutes of searching for this case yourself, and I’m not going to pretend a well-known thread doesn’t exist just because it’s uncomfortable. But I want to be exactly as careful with an uncharged, unconvicted person as I would want someone to be with me. A disputed DNA result is not a conviction. A nickname is not a name. A theory a journalist believes strongly enough to publish a book about is still, legally and morally, a theory. Sugaya Toshikazu’s story is the proof, sitting right in the middle of this same case file, of exactly what happens when certainty about DNA evidence outruns what the evidence can actually support. I am not going to make that mistake twice in the same piece.
What I can tell you plainly: police reportedly looked into this individual and did not pursue an arrest. Whether that is because the evidence genuinely didn’t hold up, or because of the kind of institutional reluctance to reopen a case that had already, once, been called solved — that I cannot tell you. Both are possible. That ambiguity is not a flaw in this piece. It is the actual, unresolved state of a twenty-nine-year-old case.
Where This Sits Now
Japan’s statute of limitations on murder was abolished in 2010 — the same reform that, in a different case entirely, let a Nagoya widower’s twenty-six-year search for his wife’s killer end in an arrest. That reform came too late for most of these five cases. Cases one through four had already passed their statutory deadlines under the old law before the 2010 change took effect, meaning that even if the true perpetrator of Fukushima’s, Hasebe’s, Oosawa’s, or Matsuda’s murder were identified with total certainty tomorrow, he could not be tried for those specific killings. Only Yukari Yokoyama’s case — the one 1996 disappearance with no recovered body, prosecutable as an ongoing abduction rather than a time-barred murder — remains legally open in the fullest sense.
Sugaya Toshikazu, for his part, received state compensation for his wrongful imprisonment and has spoken publicly, in the years since his release, about wanting the actual killer found — not for his own vindication, which the court already gave him in full, but because somewhere in Tochigi or Gunma, a person who did this to at least one child, and quite possibly several, has now lived an entire adult life in the space where a wrongly convicted man’s life should have been instead.
Say their names before you close this tab, all five, because that is the only thing left that is entirely without ambiguity in this whole case: Maya Fukushima. Yumi Hasebe. Tomoko Oosawa. Mami Matsuda. Yukari Yokoyama. The oldest of them was eight. The youngest were four. And Sugaya Toshikazu — say his name too, because he is not a footnote to this story, he is a second victim of it, a man who lost seventeen and a half years of an ordinary life to a mistake made with a microscope and absolute institutional confidence, and who walked out in 2009 into a world that had moved on completely without him.
Cold does not mean dead, and wrongly closed does not mean solved. Somewhere in a narrow stretch of land on the Tochigi–Gunma border, the actual answer to what happened to at least one of these girls — and quite possibly all of them — is still sitting, unresolved, decades after a court told an innocent man with total certainty that the science had already found him.
Sources
- 足利事件 (Ashikaga Incident) — 日本事件データベース (jiken-db.com), detailed timeline of arrest, original MCT118 DNA testing, conviction, 2009 retest methodology, and 2010 acquittal.
- 足利事件 — search-aggregated coverage citing Kotobank, official case chronology (arrest Dec 1–2 1991, indictment Dec 21 1991, conviction July 7 1993, release June 4 2009, acquittal March 26 2010).
- 清水潔『殺人犯はそこにいる — 隠蔽された北関東連続幼女誘拐殺人事件』(Shinchosha, bunko edition) — publisher’s official book description; winner of the Shincho Documentary Prize and the Mystery Writers of Japan Award.
- Reporting and reader/critical discussion of Shimizu’s investigation and the “Lupin” (ルパン) suspect theory, including the disputed DNA testing conducted via Prof. Honda Katsuhiro (University of Tsukuba) — treated in this piece as a documented but unresolved and unproven theory, not as an established fact.
- National Diet (衆議院, shugiin.go.jp) parliamentary question record referencing the case’s official status.
- 群馬県警察 (Gunma Prefectural Police), official case page for the Yukari Yokoyama abduction — case status, CCTV suspect description, ongoing appeal for information.
- Case-chronology aggregators (zikenroku.blog.jp, gobou-chan.com) cross-checked against each other for the Fukushima, Hasebe, and Oosawa case details — last confirmed sightings, discovery dates and locations, statute-of-limitations dates.
- 上毛新聞 (Jomo Shimbun), regional Gunma newspaper, coverage referencing the Tone River remains discovery.
- English-language cross-reference: Wikipedia’s “North Kanto serial kidnappings and murders” — used only to cross-check chronology, not cited as a primary source.
If you are reading this from Japan and have any information about the disappearances of Maya Fukushima, Yumi Hasebe, Tomoko Oosawa, Mami Matsuda, or Yukari Yokoyama, Tochigi and Gunma prefectural police are still accepting information. Yukari Yokoyama’s case in particular remains legally prosecutable — she has never been found, and it is not too late for that to change.