In a Tacoma restaurant on a summer afternoon in 2018, a detective sat a few tables away from a stranger, waiting for him to finish his meal. When the man wiped his mouth, crumpled his napkin, and left, the detective rose, walked to the now-empty table, and picked it up. It was sent straight to the forensic laboratory. What the analysis revealed closed a case that had remained open for 32 years.
Sixteen days later, on June 20, 2018, a Washington State Patrol unit pulled over a car during a routine traffic stop in Lakewood, Washington. Behind the wheel was a 66-year-old nurse named Gary Charles Hartman. He had never been arrested in his life for anything. No traffic violations worth mentioning, no flag in any system.
But the genetic material recovered from that napkin matched the DNA collected more than three decades earlier from the body of a 12-year-old girl named Michella Welch. Nobody in that restaurant had any idea what was happening around them. To the other customers, they were just two strangers finishing their meals on an ordinary day. No one could have imagined that one of them was quietly ending one of the most baffling murder cases in the city’s history, or that the other had spent three decades believing he had gotten away with it completely.
How does a man with a clean record for 32 years end up under secret police surveillance for the murder of a child, a killing that took place when he was still in his twenties? The answer does not begin with Hartman. It begins with Michella, and with an ordinary spring morning in a park in North Tacoma, when nobody had any reason to believe it would turn into anything other than a normal Wednesday. It was March 26, 1986.
Michella Welch was 12 years old, a sixth grader with two younger sisters whom she looked after the way older siblings do, without thinking about it. That morning, the three girls went to Point Defiance Park in North Tacoma to play. It was an extension of the family backyard for the neighborhood’s children, a green space cut through by a wooded ravine, close enough to reach on foot and familiar enough that nobody worried. Parents in North Tacoma had let their children roam that park for years without a second thought.
That was about to change forever. Tacoma already carried an uneasy weight in American true crime history long before that Wednesday in 1986. Ted Bundy had grown up a few miles from Point Defiance Park, in the same northern end of the city, decades before becoming one of the most infamous serial killers the country had ever produced. By the time Michella and her sisters were playing in that park, the city lived with a quiet, anxious awareness that violence did not always announce itself with a stranger lurking in the shadows.
Sometimes it grew up in an ordinary house on an ordinary street, indistinguishable from any other, waiting. Around ten in the morning, Michella decided to ride her bike to get lunch for herself and her sisters. It was a small, routine errand, the kind that happens in thousands of parks on thousands of ordinary days without anyone paying attention. Meanwhile, her sisters wandered toward the nearby ravine to keep playing, waiting for her to return with the sandwiches.
No one noticed her leave with any sense that something was wrong. Why would they? Children rode their bikes through that neighborhood all the time. Michella never came back.
When her sisters returned to the spot where they had agreed to meet, Michella was gone. Her bike was still there, and so were the sandwiches she had gone to fetch, as if she had simply vanished. Whatever happened to her happened fast, and close enough that a girl bringing lunch to her siblings did not return to finish the task. What followed was a frantic search.
At first, family members called her name through the trees. Then a full search-and-rescue unit combed the wooded area around the park. They found her in an isolated section of the ravine, more than a quarter of a mile from the playground where her sisters had been waiting. The Pierce County medical examiner determined she had died from a severe head injury.
The examination also revealed evidence that she had been sexually assaulted before her death. A 12-year-old girl had been killed in broad daylight in a public park, within earshot of where her sisters were playing. The perpetrator vanished without leaving a clear name, a clear motive, or a clear trace. Tacoma in 1986 had no answer to the most important question: who could do this to a child and then simply return to an ordinary life as if nothing had happened?
Word of what happened spread through Tacoma quickly. Parents who had let their children roam the neighborhood parks for generations suddenly found themselves walking them to school, watching them from front porches, locking doors that had never needed locking before. A city built on the assumption that its parks were safe had been shown, in the most brutal way possible, that the assumption was wrong. And not knowing who did it was a particular kind of torment.
There was no face to be angry at, no name to wait for—just a void where an explanation should have been. Detectives worked the case hard in the weeks and months after Michella’s death. They interviewed neighbors, swept the park, and followed up on anyone who matched even a general description of a man seen near the ravine that morning. Tips came in.
Leads were chased. Suspects were examined and cleared, one after another. But 1986 was a different world for forensic science. DNA testing, as we understand it today, was rare and largely unusable.
There was no database to run a DNA profile against, even if investigators had been able to collect one. The tools that would eventually solve this case were still decades away from being invented, locked in the future, useless to a detective standing in that ravine trying to make sense of what he was seeing. Some of the original investigators who worked Michella’s case in 1986 retired without ever seeing it solved, handing over the case files, the evidence logs, and their own theories to whoever inherited the file next. That is how most cold cases actually move through a department over the decades: not through one continuous investigation, but through a chain of detectives, each picking up where the last one left off, each hoping to be the one who finally made progress.
The previous generation had no access to a confession, and without a witness who had seen something concrete, the investigation slowed. Then it slowed more. Then, like so many cases from that era, it went cold. Michella’s family kept living in the same city where her killer, whoever he was, presumably still lived too.
That is the particular cruelty of an unsolved murder in a mid-sized city. The person responsible is not a faceless monster in a distant place. He is a neighbor, a coworker, someone standing behind you in the grocery line, and nobody knows it. The years passed the way they pass when a case goes quiet—not all at once, but through the slow accumulation of ordinary days.
Michella’s sisters, the two girls who had waited for her at the ravine that morning, grew up without her. There were birthdays, graduations, milestones she never saw and was not there to share. The physical evidence collected during her autopsy remained carefully preserved in storage, waiting for a technology no one could have predicted, while the city around it moved on without her. The case sat closed for 20 years before science caught up even a little.
In 2006, Tacoma police returned to the physical evidence collected during Michella’s original autopsy and extracted a male genetic profile from swabs taken at the time. It was a real step forward, a breakthrough that brought renewed hope to a family after decades of silence. Somewhere within that genetic file was the signature of the person responsible. The problem was that a DNA profile only helps if there is something to compare it to.
When investigators searched state and national databases, they found no match. Whoever this man was, he had never been arrested for a crime serious enough to put his DNA on record anywhere in the country. The trail went dead again, this time for another decade. The genetic file sat in storage, intact and useless at the same time.
The thread that finally moved this case forward did not come from a new witness or a deathbed confession. It came from genetic genealogy, and from a woman who had spent years building family trees for a living before she started using those same skills to track down killers. Raye Winter entered this work indirectly. She trained as a biologist, then spent years as a patent attorney specializing in biotechnology.
It was the kind of career that required reading complex scientific claims for a living, understanding precisely what they proved and what they did not. At some point, she turned her attention to tracing her own family history as a personal hobby, the way millions of ordinary people do in their evenings and weekends. That hobby became expertise, and that expertise became a tool law enforcement had never had before. A retired patent attorney with a gift for mapping out families, capable of doing what no crime lab could accomplish on its own.
To understand how it worked, it helps to understand what had changed. Traditional forensic DNA testing only works if the suspect’s genetic profile is already in a criminal database, waiting to be matched against crime scene evidence. If he was never arrested, never swabbed, had no reason to be in that system, the match would never happen, no matter how many decades passed. Investigative genetic genealogy turned that situation upside down.
Instead of searching for the suspect directly, genealogists took the crime scene DNA and searched for his relatives—people who shared pieces of his genetic code because they shared ancestors. Second cousins, distant cousins, people who had uploaded their DNA to a genealogy website out of curiosity about their family origins, with no idea that a distant relative of theirs had committed a murder decades earlier. From those partial matches, a genealogist could build a family tree, trace it back through generations of history, then work forward, narrowing a list of thousands of possible relatives down to a single family. The technique had made national headlines two months earlier, when investigators used it to identify a serial killer and rapist who had terrorized California for decades in a case that had been gathering dust since the 1970s.
Barbara Rae-Venter was the genealogist behind that achievement as well. By the time Pierce County contacted her, she had already proven that no cold case, no matter how old, was out of reach. In February 2016, and over the following year, the Pierce County Prosecutor’s Office made a decision that changed the course of Michella’s case completely. They brought in Parabon NanoLabs, based in Virginia, along with Rae-Venter herself, to see whether the old crime scene DNA could be used more deeply than a standard database search would allow.
Rae-Venter took the genetic material from the Michella case and created a detailed profile from it, designed to capture far more information than a standard law enforcement database would look for. That profile was uploaded to two public genealogy platforms, FamilyTreeDNA and GEDmatch—the same kinds of websites ordinary people use to find long-lost cousins or trace their ancestors’ origins. In this case, the goal was not reunion. It was identification.
None of this was without controversy. As police departments across the country began uploading crime scene DNA to public genealogy sites, questions arose about consent, about whether someone who uploaded their own genetic file just to learn about their ancestry should have any say in whether that same data would one day help identify a killer among their distant relatives. GEDmatch later changed its policies to require users to explicitly opt in before law enforcement could search their information for violent crime cases. But in 2016 and 2017, while Rae-Venter was quietly assembling the family tree that would eventually lead to Hartman, those debates were just beginning, and Michella Welch’s case became one of the earliest examples to draw on.
From that point on, the work looked less like a police investigation and more like painstaking historical research. Rae-Venter and her team pored over public genealogy records, census documents, and old obituaries to build a family tree rooted in the partial genetic matches the uploaded profile had revealed. Slowly, generation by generation, the tree narrowed. Each relative added to the chart either widened the search or closed off an entire branch.
It was slow, meticulous work. By 2017, the results had narrowed completely to a specific family that had lived in the northern end of Tacoma in 1986, in the same neighborhood surrounding Point Defiance Park. That family included two brothers, both of whom had lived in the same area of Tacoma at the time of the crime. Their geographic and generational profiles matched what the genealogists had deduced from the family tree.
Separately, both the Parabon team and Rae-Venter’s independent analysis pointed to the same two men as the most likely candidates. One of those brothers was named Gary Charles Hartman. Having two names instead of one narrowed the search considerably, but it also prevented investigators from simply arresting the first name on the list. Family trees identify families, not individuals.
Although the tree pointed to two brothers, it left open the question of which one, if either, had been in that ravine in 1986. That distinction could only be settled by a direct DNA comparison, which meant the investigation still had one more difficult, delicate step before anyone could be arrested. Genetic genealogy could guide investigators to a family, even narrow that family to a handful of specific people. But what it could not do, at least not on its own, was put a suspect’s name on an arrest warrant.
Investigators needed something they could actually use in court: a direct DNA sample they could compare against the 1986 evidence without alerting Hartman that he was a suspect. So they watched him. Beginning on June 4, 2018, Tacoma detectives began quietly surveilling Gary Hartman, tracking his movements, waiting for an opportunity that would not tip him off. He was at the time an ordinary 66-year-old man living an ordinary life, working as a nurse at Western State Hospital, caring for patients day after day, completely unaware that a team of investigators believed they knew what he had done more than three decades earlier.
Imagine that for a moment. A man who had spent 32 years going to work, paying his bills, living quietly in the same area where he had killed a child, with no one around him suspecting a thing. During those two-plus weeks of surveillance, investigators had to walk a razor’s edge. If they moved too fast, they risked exposing Hartman before they had any conclusive evidence to confront him with in court.
If they moved too slow, the man they believed had killed a 12-year-old girl remained free, going to work at the hospital, living his ordinary life, completely unaware that he was being watched. Every day of that surveillance was quiet, unglamorous pressure—the kind that never makes headlines but determines whether a decades-old case gets solved or slips away again. The opportunity came at a restaurant. A detective watched Hartman eat, watched him wipe his mouth and set down the napkin.
After he left, the detective collected that discarded napkin as what is known as abandoned DNA—genetic material a person leaves in a public place in which they no longer have any legal expectation of privacy. No warrant was needed for a napkin thrown away in a public restaurant, because courts have long treated items abandoned in public spaces as available to anyone. Legally, it was no different from trash left on the curb. The method has since become nearly routine in cold case investigations: detectives quietly follow a suspect through an ordinary day, waiting for a coffee cup, a cigarette butt, a used fork at a restaurant—anything carrying genetic material the person discarded voluntarily.
The napkin was sent immediately to the Washington State Patrol crime lab for testing. On June 20, 2018, the lab results came back. The DNA on that napkin matched the DNA extracted from the crime scene in 1986. After 32 years, investigators no longer had just a family tree and a theory.
They had a direct match. Hartman was arrested the same day during a traffic stop in Lakewood, taken into custody without resistance. A 66-year-old man with no criminal record at all was suddenly facing the weight of a murder committed when he was still in his twenties. Two days later, on June 22, Pierce County Prosecutor Mark Lindquist filed formal charges against Hartman: first-degree murder and first-degree rape in the death of Michella Welch.
Bail was set at $5 million, a figure that made clear just how seriously the court treated a case that had waited more than three decades to reach this moment. News spread through Tacoma like wildfire. For 32 years, residents had lived under a question mark, wondering privately whether the person responsible for Michella’s death was still free, still walking the same streets, perhaps still living in the same neighborhood. Now there was a face and a name—and it belonged to someone no one would have picked out based on reputation alone.
A quiet nurse who had spent decades caring for vulnerable patients in a state hospital. Someone whose adult life, on the surface, showed no visible sign of what he was capable of. That contradiction—a caregiver by profession and a child killer by his past—was harder for people to absorb than a stranger would have been. Years of legal proceedings followed before the case finally reached its resolution.
Hartman eventually faced a bench trial, meaning the case was decided by a judge instead of a jury of his peers. The evidence rested on the same science that had taken 32 years to arrive at. The match between the DNA on the discarded napkin and the genetic profile extracted from the 1986 crime scene was presented systematically to a single judge rather than twelve jurors. The hearing itself was remarkably short given everything that had come before it, lasting no more than two hours before the judge delivered his verdict.
A very short time to bring closure to a case that had taken 32 years just to reach the courtroom. Gary Charles Hartman was convicted of first-degree murder in the death of Michella Welch. At sentencing, Judge Stanley Rumbaugh imposed a sentence of 26 years and six months. Hartman, then 70 years old, faced a punishment that would keep him behind bars for the rest of his life.
A crime committed in his twenties had finally caught up with him as he entered his seventies, closing a gap of more than three decades between the commission of the crime and accountability for it. For Michella’s family, the sentencing represented something they may have stopped expecting. Decades of living with an unanswered question, watching other families in Tacoma get their own resolutions while their case remained open forever, ended with a single moment in a courtroom, when the judge finally spoke the words that made it official. It did not erase 32 years.
Nothing could. But it closed the door on the mystery, which may be the closest thing a case can offer a family in their position to peace. Michella Welch’s case did not remain isolated once it broke. It turned out to be the first of two murders of young girls in Tacoma in 1986 that were eventually solved through investigative genetic genealogy.
The second was Jennifer Bastian, 13 years old, killed the same year under circumstances so similar to Michella’s that investigators quietly wondered for decades whether the same person was responsible for both. Genetic evidence linked Gary Hartman to Jennifer’s death as well, connecting two cold cases that had haunted the same city in the same year, and in some cases, the same investigators, for more than three decades. Two families who had spent their lives wondering finally got the same answer within the same investigation, from a discarded napkin. For anyone in Tacoma who lived through 1986 as a parent, the resolution of both cases at the same time was not an end to the tragedy so much as a closing of a door.
Two girls killed months apart in the same city. Both cases had weighed on grieving families for more than three decades. Both were answered at last because one genealogist had been willing to spend years tracing the ancestry of strangers, hoping that one of those branches would lead somewhere significant. For 32 years, the man responsible for Michella Welch’s death lived an ordinary life.
He worked a steady job caring for patients in a state hospital. He had no criminal record worth mentioning, no prior arrest that would have put his DNA in a database, nothing that would have identified him to investigators through traditional means. The case was not solved because of a mistake he made in 1986. It was solved because the science of identifying people through their extended family trees had finally caught up with a single trace of genetic material left at the crime scene, waiting all that time to be revealed.
What is worth reflecting on is how much this has affected every other unsolved crime sitting in archives. For decades, a murder with DNA evidence but no database match was a dead end. The evidence existed, but there was nothing to point it at. Investigative genetic genealogy took that same evidence and aimed it down an entirely new path—not toward a name in a criminal database, but toward a family tree built by ordinary people who had no idea their curiosity about their ancestry would one day help catch a killer.
In the years since Michella Welch’s case broke, the same approach has been used to identify suspects and victims in cold cases across the country. Murders, sexual assaults, unidentified remains that had gone unsolved for decades, with no other apparent avenue for resolution. What started as a small group of genealogists working case by case has become a standard tool that police departments turn to whenever old DNA evidence exists but databases offer nothing to compare it to. None of it undoes what happened to the victims.
It only means that families waiting for answers no longer have to accept the possibility that there are none. Somewhere right now, in an evidence locker in another city, there is another sample from a crime scene in another case from decades ago, waiting for a genealogist willing to spend months tracing the ancestry of strangers, hoping that one of its branches leads somewhere that matters. Michella Welch’s case is proof that the burden does not have to be permanent, even when it lasts 32 years. Michella Welch was 12 years old.
She rode her bike to get lunch for her sisters on a Wednesday morning in March, and she never came home. It took 32 years, a discarded napkin, and a genealogist tracing a family tree through census records and old obituaries to finally answer the question that had haunted her family since 1986. Some cold cases stay cold because there is no evidence at all.
This one stayed cold only because the tools needed to solve it had not been invented yet.
