For years, the worst comfort a killer could lean on was time.
In Cedar Rapids, that passage of time could be measured against a black dress. Michelle Martinko had worn it to a choir banquet on December 19, 1979. Police recovered it after finding the eighteen-year-old dead in her family’s car. They preserved the bloodstained fabric, but the testing available then could not tell them everything it held.
Decades later, a stain on that dress led investigators to a family, then three brothers, then a drinking straw left behind in a restaurant. The evidence had not arrived late. The means of understanding it had.
This collection follows that change through cases in which old evidence produced new suspects, trials or identities. The outcomes differ: a conviction is not the same as a police identification after a suspect’s death, and recovering a victim’s name does not necessarily solve the murder. Those distinctions are part of the story.
When Time Stops Protecting the Guilty
Sometimes the answer comes from a technology that did not exist when the crime happened. Sometimes it comes from evidence that police had all along but could not fully use. The cases below show both possibilities, alongside the interviews and courtroom scrutiny that still have to follow a laboratory result.
Cases Broken Open by DNA and Genetic Genealogy
Joseph James DeAngelo
For decades, the crimes associated with the Golden State Killer crossed jurisdictions and accumulated different names: the Visalia Ransacker, East Area Rapist and Original Night Stalker. The man eventually prosecuted for them had himself worked as a police officer. The distance between his public employment and his crimes made the identification particularly disturbing.
Investigative genetic genealogy led authorities to Joseph James DeAngelo in 2018. In June 2020, he pleaded guilty to thirteen murders and thirteen kidnappings to commit robbery, also admitting numerous uncharged crimes. Prosecutors said the ages of victims and witnesses, the deaths of other witnesses and the prospect of a lengthy prosecution informed the plea agreement.
The hearings took place in a university ballroom so victims and families could attend while observing pandemic distancing requirements. In August 2020, DeAngelo received life sentences without parole. Scientific identification had made prosecution possible; the proceedings then required him to answer for particular people and particular crimes.
Read the full story: The Officer and the Monster: The Golden State Killer’s Double Life.
Jay Cook and Tanya Van Cuylenborg
Jay Cook and Tanya Van Cuylenborg crossed from British Columbia into Washington in November 1987 to collect furnace parts for Cook’s father. Their plan was practical: reach Seattle, sleep in the family van, buy the parts and go home. Instead, their bodies were found in separate counties. The van turned up in a third.
Biological evidence recovered from Van Cuylenborg was preserved. In 2018, genealogical research pointed toward William Earl Talbott II. Police then collected a cup he had used, and the state laboratory compared that sample with the crime-scene DNA. A possible family relationship had become a directly testable lead.
But identifying the source of DNA was not the whole murder case. At trial, the evidence also included a palm print on the van, the circumstances of the couple’s journey, the locations of the bodies and Talbott’s familiarity with the area. Talbott disputed that the evidence proved he killed them. The appellate record acknowledges gaps and conflicting evidence, including weaknesses in some of the prosecution’s inferences about sexual contact.
The Washington Court of Appeals nevertheless affirmed the two aggravated-murder convictions in its February 2024 opinion, while directing correction of an erroneous firearm enhancement. The state Supreme Court denied further review that July. The case demonstrates the reach of genealogy without turning a family-tree lead into a substitute for a trial.
Two RTRS articles cover this same investigation: Jay Cook and Tanya Van Cuylenborg’s journey and William Talbott II: the evidence and appeals.
Michelle Martinko
Michelle Martinko’s murder haunted Cedar Rapids for nearly forty years. After the choir banquet, she drove to Westdale Mall, spoke with friends and left the building that evening. Police found her in the Buick in the early hours of the next morning. Her dress became one of the objects through which investigators kept returning to that night.
Progress came in stages. Early DNA testing could not isolate the answer investigators needed. In 2005, another stain on the dress yielded a partial male profile. A search of the criminal DNA database produced no match. More than a hundred possible suspects were excluded through comparisons, yet the man whose DNA mattered remained unidentified.
Genealogy changed the search in 2018. Researchers worked backward through family relationships, and a cooperating relative’s sample narrowed the inquiry to three brothers. Samples from discarded items eliminated two. Police followed Jerry Burns to a restaurant, watched him drink through a straw and collected it after he left.
The straw result did not exclude Burns as the contributor. Investigators then obtained a warrant for a direct mouth swab, which confirmed a match to the profile from the dress. That sequence matters: the family tree suggested where to look, and further testing supplied the comparison.
A jury convicted Burns in 2020, and he received life without parole. The Iowa Supreme Court affirmed in 2023. In May 2026, KCRG reported that the state was seeking dismissal of his post-conviction claims; that report described a request, not a new ruling overturning the conviction. The later challenge does not erase the original investigative path, but it belongs alongside the account of the breakthrough.
Read the full story: Michelle Martinko: the evidence, conviction and later challenge.
Rita Curran
A cigarette butt collected near Rita Curran’s body in 1971 became crucial evidence. The twenty-four-year-old teacher had been killed in her Burlington apartment. DNA testing in 2014 produced no database match.
Genealogy later pointed to her upstairs neighbor, William DeRoos. Police said DNA on Curran’s jacket also matched. His former wife then acknowledged that he had gone out that evening and asked her to lie about it, contradicting their original account.
Police closed the case in February 2023. DeRoos had died in 1986. At the announcement, Curran’s brother spoke about Rita, their parents and what the family had endured. The focus returned to the people whose lives the murder had changed.
Read the full story: Rita Curran and the preserved evidence.
Milestones in These Investigations
- 1971, 1979 and 1987: Evidence was collected in the Curran, Martinko, and Cook–Van Cuylenborg investigations respectively.
- 2018: Genetic genealogy helped identify DeAngelo and Talbott; family research and subsequent testing also brought the Martinko investigation to Burns.
- 2020: DeAngelo pleaded guilty, and Burns was convicted. Both received life sentences without parole.
- 2023: Burlington police identified Curran’s killer; the Iowa Supreme Court affirmed Burns’s conviction.
- 2024: Talbott’s convictions were affirmed, and the Washington Supreme Court denied further review.
- 2026: Separate identification efforts restored Melinda Beardsley’s name and identified the remaining Martin family members. Their outcomes are explained below.
What Actually Breaks a Cold Case Open?
The evidence in these cases did not all become useful at once. Martinko’s dress went through successive rounds of testing. Curran’s cigarette yielded a profile before it yielded a name. Talbott’s identification needed a comparison sample and a wider evidentiary case.
Genetic genealogy looks for relatives of an unknown DNA contributor and uses family records to develop candidates. Investigators then have to establish whether the candidate’s DNA matches and what that connection means. A match can connect someone to evidence; the rest of the record must address when, how and whether it proves the crime charged.
Not every breakthrough begins with a family tree. In December 2021, Texas authorities reported that more than sixty of Samuel Little’s ninety-three murder confessions had been matched to victims through DNA evidence or extensively corroborated interviews. Ranger James Holland’s interviews supplied leads, but investigators still had to test them against cases across jurisdictions. Authorities warned that Little could be badly wrong about dates and distances. Confession alone could not safely supply a complete chronology.
Read more: Samuel Little and the work of identifying his victims.
Beyond Murder: Cases Where Identity Was Restored
The Martin family’s disappearance had remained unresolved since December 1958. A diver located their car in 2024; recovery work and later dives produced components and human remains. In April 2026, authorities announced the identification of Kenneth, Barbara and their daughter Barbara, known as Barbie. A DNA profile for Kenneth, considered alongside the other evidence, supported the identifications. The Hood River County Sheriff’s Office closed the investigation after finding no evidence of criminal activity. Read the Martin family investigation.
The result was different for Melinda “Pip” Beardsley. Her body had been found in an Emigrant Gap snowbank in 1977. Research by Moxxy’s Missing Persons Task Force connected her disappearance with the unidentified victim; a family DNA comparison confirmed her identity in February 2026. The March announcement said the homicide investigation continued. Restoring her name answered who she was, while opening a clearer route toward investigating who killed her.
An Unsolved Case Belongs in a Different Category
Dorothy Jane Scott’s case remains an important comparison, but not an example of a solved murder. The Orange County Sheriff’s Department’s case page says the mysterious caller regarded as a potential suspect was never identified or found, and investigators continue to seek leads. Her remains were identified after their discovery in 1984. That identification must not be mistaken for the identification of her killer.
What an Answer Can—and Cannot—Return
The most consequential difference between these stories lies in what became possible afterward. DeAngelo faced the people whose lives his crimes had changed. Burns and Talbott faced trials and appellate review. Curran’s family received a police finding without the possibility of a trial. Beardsley’s family recovered her name while the homicide inquiry continued.
None of those outcomes gives back the years. But each replaces a particular absence with something that can be examined: a name, a tested connection, a judgment, or a recovered person. The black dress police kept after Michelle Martinko’s last evening could not bring her home. Preserving it meant that her last evening did not have to remain beyond investigation.
Frequently Asked Questions
How do cold cases get solved after decades?
These cases combined preserved evidence with later testing, genealogy, interviews and renewed review. Different evidence moved each investigation forward. A family-tree lead could supply a name for direct testing. A recorded confession needed independent corroboration. None provides a guarantee that every old case can be resolved.
Why does DNA matter so much in cold cases?
A usable biological sample can be tested long after the event. In Martinko’s case, improved testing revealed a male profile in preserved evidence, and later family research supplied candidates for comparison.
What is genetic genealogy in criminal cases?
It combines genetic relationships with family-history research to develop leads about an unknown person. It does not replace direct comparisons, investigation or the legal burden of proving guilt.
Can police close a case if the suspect is dead?
Yes. Burlington police closed Rita Curran’s case after identifying DeRoos. That was an investigative finding, not a conviction: his death prevented prosecution.
Does identifying a victim solve the homicide?
No. Melinda Beardsley’s identification and the investigation of her killing are separate achievements. Similarly, Dorothy Jane Scott’s identification did not establish who killed her. A responsible account should say precisely which question has been answered.
Research Sources
The following resources were consulted to verify and update this article.
- Orange County District Attorney: DeAngelo’s June 2020 guilty pleas — Documents DeAngelo’s guilty pleas and the counts to which he admitted guilt.
- Sacramento County DA’s January 2021 review — Records sentencing and arrangements for the hearing.
- Snohomish County: the 2018 Talbott arrest and DNA investigation — Describes the identification and arrest of William Talbott II.
- Washington Court of Appeals, February 2024 opinion — Explains the appellate assessment of evidence and disposition, including the firearm enhancement.
- July 2024 review results — Records the denial of further review in July 2024.
- Iowa Supreme Court: State v. Burns, March 2023 — Details the Burns investigation, straw comparison, conviction and appeal.
- 2020 sentence — Reports Burns’s life sentence without parole.
- state’s May 2026 response to post-conviction claims — Documents the state’s May 2026 request concerning post-conviction relief.
- Vermont Public: February 2023 Curran press conference — Reports Burlington police’s identification of William DeRoos and the physical evidence.
- Texas DPS: December 2021 Samuel Little investigation update — Distinguishes cases corroborated across jurisdictions from unmatched confessions.
- Multnomah County Sheriff’s Office: Martin family findings — Gives the identification of the Martin family and the reason the inquiry closed.
- Moxxy Forensic Investigations’ March 2026 Beardsley announcement, published by YubaNet — Reports the identification of Beardsley and the unresolved homicide inquiry.
- Orange County Sheriff’s Department: Dorothy Jane Scott — Records the condition of the Scott investigation and the request for information.





