Three separate law enforcement agencies handled the same piece of evidence over the course of 36 years. The FBI reviewed it. County investigators reviewed it. Sheriff’s detectives reviewed it more than once.
None of them noticed what was sitting directly in front of them. For nearly five decades, the murder of a young woman in Snohomish County, Washington remained unsolved—not because the evidence was missing, but because it had been there all along, waiting in an evidence room for someone to examine it under the right conditions. Three agencies had the opportunity. All three missed it.
On August 23, 1972, Judy Loomis was 20 years old. She lived on Winesap Road with her parents, John and Rosemary, her fiancé, Jim Roberts, and her 12-year-old sister. She kept a horse named Saudi in a stable about six miles from home and rode with an English saddle whenever she visited. People who knew her described her the same way over and over: blunt and direct, the kind of person who would pick up injured animals from the side of the road and bring them home.
She had painted a mural of running horses on her bedroom wall, floor to ceiling, and was planning her wedding. Most days, her parents drove her to the stable. That afternoon, she decided to make the trip alone for the first time. Before leaving, she took a pair of boots from her sister’s room without asking—brown hiking boots with waffle soles, size five and a half.
She grabbed Saudi’s bridle, got on a white ten-speed bicycle, and headed down the driveway. A neighbor named Linda Hock, who was home that day packing to move out of state, watched her leave from a back window. It was nothing more than a wave between two people close in age. Judy rode north and east through open farmland.
About three and a half miles later, she passed a roadside fruit stand near the Everett Highway where 14-year-old Kenda Mathieu was working that afternoon. Mathieu noticed her for two reasons: she was riding a bicycle while wearing boots, which seemed unusual, and she paused at the intersection longer than the situation required—long enough that Mathieu remembered it decades later—before crossing and turning onto a narrow, tree-lined road called Penny Creek Road. Then she disappeared into the trees. It was later confirmed that Kenda Mathieu was the last person to see Judy Loomis alive.
One detail from that evening was never resolved. Judy was carrying her horse’s bridle when she left home. It was not with her when she was found. It was never recovered—not that day, and not during nearly 50 years of investigations that followed.
To this day, no one knows where it went. Shortly after 5:30 that evening, a young couple drove down the same dirt track to shoot cans in the woods, a common way to spend a summer evening in that area in 1972. A fallen tree blocked the road. When they got out to move it, they found Judy near it, badly injured.
They carried her to their car and drove west toward Edmonds to Stevens Memorial Hospital, the building known today as Swedish Edmonds. By the time they arrived, she was dead. Investigators reconstructed the attack within a time frame between 5:00 and 5:30 that evening. Judy had been forced off the road at gunpoint, sexually assaulted, and shot once with a small-caliber handgun.
Her bicycle was found near her body. What investigators did not have was almost everything else. No fingerprints. No cartridge casing linked to a weapon.
No one heard or saw the attack itself, even though it occurred near a road people used regularly. They built their case around who had actually been in the area. A local landowner had a history that worried investigators. Neighbors had noticed his behavior around Judy before, though no one ever filed a formal complaint.
He was questioned and released. No physical evidence connected him to the crime scene. A man cutting wood near the dirt track that evening could not clearly account for his hours. He was questioned and eventually released.
Investigators also turned to a motorcycle club known to pass through the area that summer, building DNA profiles of member after member. Each one was cleared in turn. By early 1973, deputies had searched nearly every farm within miles of Penny Creek Road, chasing any report of an unfamiliar car or stranger near the woods that evening. Target shooting in rural areas was common there, which meant the single important bullet faded into an environment where gunfire was just background noise.
None of it led anywhere. By the end of 1972, the case was set aside as unsolved. The evidence—Judy’s clothes and the boots she had borrowed—was stored in the Snohomish County Sheriff’s Office, where it remained untouched for more than three decades. John and Rosemary Loomis spent the rest of their lives without an answer.
Both died still waiting for a name, an arrest, or anything that explained why their daughter never came home. Neither lived to see the case solved. The burden fell to Judy’s younger sister, who grew up and became publicly known as Jana Loomis Smith. She called the sheriff’s office regularly for decades, just to ask if there was anything new.
Most of the time, there wasn’t. She kept calling anyway. She attended press conferences unrelated to the case just to remind the department that her sister’s file was still open. The department did not let the case fade entirely.
In 2008, Judy’s case was printed on a set of cold case playing cards distributed through the county jail system, hoping a small detail would reach the right person. Chuck Wright, a volunteer at the sheriff’s office, carried one of those cards in his wallet for the next ten years and wrote about Judy periodically in a local column, later saying some cases stay stuck in a person’s mind. None of it produced a name. In 2005, detective Jim Scharf was assigned to the county’s cold case unit and pulled Judy’s file, by then the oldest unsolved homicide on record.
He reconnected with anyone still reachable and sent the remaining evidence for examination using methods that did not exist in 1972. Three years later, a forensic scientist at the Washington State Patrol crime lab reexamined the boots under newer lighting and equipment. On the outside of the left boot, she found something every previous examination had missed: a small stain. It had been there since 1972.
It had passed through the medical examiner’s office, sheriff’s investigators, and even a previous FBI review. The equipment needed to find it had existed since the late 1980s. What had concealed it was not a limitation of forensic science, but simply the chain of custody—36 years of handling, evidence lockers, and transfers between agencies that had never been fully traced. A partial DNA profile was extracted from the stain and entered into CODIS, the national database that compares crime-scene DNA to people convicted of qualifying offenses.
It matched nothing. The profile sat in that database for the next ten years without a single hit—not because the offender had no record with the law, but because his history did not legally require him to submit a sample. The discovery that should have closed the case instead caused another decade of waiting. By 2018, investigators had a confirmed profile, a precise timeline, and a list of suspects cleared long ago.
What they still lacked was a match. That year, Jim Scharf tried something the case had never attempted. He sent the profile to Parabon NanoLabs for forensic genetic genealogy analysis, a technique that had entered public awareness only months earlier through the identification of the Golden State Killer in California. CODIS only works if the responsible person has a recorded conviction.
Genetic genealogy asks who the person is related to. The profile is compared against public genealogy databases like GEDmatch, where ordinary people had voluntarily shared their DNA for family history research. Any relative in that system, even a distant one, creates a thread an investigator can trace backward, generation by generation. Genetic genealogist Deb Stone spent 57 hours building a family tree from that thread, narrowing the unknown man to six brothers, all raised within a few miles of where Judy disappeared.
Investigators quietly eliminated five of them using employment records and known timelines without alerting the family. Only one man remained. A name from a family tree is not evidence of guilt. It is only a starting point.
Investigators still needed a direct DNA sample from him, collected without his knowledge in a public place. On August 28, 2018, a surveillance team followed him to the Tulalip Resort Casino north of Everett. He ordered a coffee, finished it, and dropped the empty cup in a trash can as he left, unaware he had just handed investigators the one piece of evidence this case had lacked for 46 years. Officers recovered the cup.
The lab ran the DNA. It matched the 2008 stain. The odds of a coincidental match were one in 980 million. The DNA belonged to a man named Terrence Miller, 77 years old at the time of his arrest and 30 in August 1972.
A lifelong resident of Snohomish County, he had worked as a heavy equipment operator with a local union from 1958 to 1995. By the time investigators reached him, he ran a small pottery business from his home garage with his wife, his partner for more than four decades. In the neighborhood, he was just Terry, a retired man who bowled and golfed and stayed active on social media until the morning of his arrest. Court records told a different story behind that retirement.
He married at 18 to a 14-year-old girl, married twice more before turning 30, and had a 1968 summons for indecent acts before a teenager, with other accusations in 1978, twice in 1990, and again in 1999—none resulting in conviction. He separately admitted to police that he had molested a young girl in the mid-1970s. But none of this history connected him specifically to Judy’s case. What investigators could establish was geographic proximity: he had lived about five miles from where she was killed and four blocks from the intersection where she was last seen, for nearly five decades, without his name ever appearing on the suspect list once.
On April 11, 2019, investigators arrived at his home and took him into custody. He was charged with first-degree murder, 47 years after Judy Loomis died on that dirt track outside Edmonds. Miller’s trial began in October 2020 before Judge David Kurtz. By then, it was the longest-awaited trial in the county’s history.
Deputy prosecutors Craig Matts and Bob Langen built their case almost entirely around the DNA evidence chain. One judicial ruling shaped the entire trial: Miller’s documented history was kept completely from the jury, deemed too prejudicial to admit. The prosecution’s case had to stand or fall on DNA alone. Defense attorneys Laura Martin and Fred Moli challenged that chain at nearly every point—decades of storage, outdated handling methods from the 1970s, and questions about possible sample contamination in the lab.
Martin argued that a DNA match was not proof of murder, describing the case as an attempt to fit puzzle pieces together that did not match. Jurors also heard a recorded jail phone call in which Miller told his wife the case could not be won. Jana Loomis Smith took the witness stand and identified the boot she had not seen since she was 12 years old, telling jurors that accountability was the only hope left for her family. Both sides called forensic experts to argue over the reliability of DNA after decades of storage, testimony that consumed several days of the two-week trial.
A jury of seven women and five men began deliberations as soon as testimony ended. Closing arguments finished on Friday, November 6, 2020. By Monday morning, Terrence Miller was dead at his home—suicide, according to the medical examiner, just hours before the jury was scheduled to resume deliberations. The jury was never told.
They deliberated that morning without knowing the man on trial was already gone, and returned their verdict that evening: guilty of first-degree murder. With no living defendant, a traditional sentence was impossible. On December 17, 2020, Judge Kurtz held a hearing and signed an order formally confirming Miller’s conviction beyond a doubt, the closest this case came to a sentencing. Jana Loomis Smith spoke, along with Richard Derry, Judy’s adopted brother, who described what five decades of waiting had cost the family.
Miller’s niece, the only family member present, gave a brief statement expressing sympathy for the Loomis family. The defense’s motion to vacate the conviction was denied. Not every question here found an answer. Judy’s horse bridle was never found.
The couple who found her that evening in 1972 were never publicly identified. Three separate agencies had direct access to that boot over 36 years. None of them saw the stain that ultimately solved the case—not because it was hidden, but because no one examined it under the right conditions until 2008. The resolution this case achieved took 46 years to arrive, not because the evidence was complicated, but because the tools needed to read it did not yet exist.
The answer was there the entire time. It simply took that long for someone to finally notice it.
