Ted Bundy and Two New 1974 Victims

Ted Bundy, Fifty-Two Years Later

Ted Bundy was executed in Florida on January 24, 1989. Thirty-seven years later, Utah authorities are still adding names to the list of people he killed.

Melissa Ann Smith was 17, from Midvale, and the daughter of the Midvale police chief. She was last seen on the night of Friday, October 18, 1974, after leaving a friend at a pizza parlor. A deer hunter found her body nine days later in Summit Park, near Park City. Laura Ann Aime was also 17. She disappeared on Halloween night that year after leaving a party in Lehi, and hikers found her the following month off American Fork Canyon Road.

Investigators announced in April 2026 that Bundy killed Aime. Then on October 27, 2026, fifty-two years to the day after Smith was found, Summit County Sheriff Kacey Bates announced that her case was closed too. Standing with him were the Salt Lake County and Utah County sheriffs’ offices, the Summit County Attorney’s Office and the Utah State Crime Lab.

Part of what made this possible is recent. Utah’s crime laboratory began using new genotyping technology in 2023 that can build usable profiles from samples that are degraded or that contain DNA from more than one person. Evidence that had sat unreadable for decades became readable.

Summit County Attorney Margaret Olson said: “This DNA evidence, when considered alongside evidence related to pattern, motive, means and opportunity, establishes that Ted Bundy was responsible for Melissa Ann Smith’s murder.” Detective Ben Pender of the Salt Lake County Sheriff’s Office, who leads its cold case unit, added that closing the case “does not close the book on Ted Bundy.” Nancy Wilcox, 16, who disappeared sixteen days earlier and whom Bundy confessed to killing, is still open, and there is no body. So are the cases of Susan Curtis, 15, who vanished in 1975 from a youth conference at Brigham Young University, and Deborah Smith, found near the Salt Lake City airport in 1976. Bundy confessed to at least thirty killings. In Utah he is believed to have killed at least five, and possibly eight.

One person who worked the case for four decades did not see it closed. Lieutenant Alan Siddoway of the Summit County Sheriff’s Office died on October 11, 2026, sixteen days before the announcement, at 62. The sheriff closed the news conference with a line Siddoway used, from Voltaire: to the living we owe respect, but to the dead we owe only the truth. The Utah County sergeant dedicated the closure to him.

Families waited two generations. The work is worth doing and the people who did it deserve credit for doing it.

The rest of this post is about the science underneath, because the way it has been reported has flattened a distinction that matters enormously in cases where the suspect is alive and the trial has not happened yet.

What the Testing Actually Showed

There are two different kinds of DNA result here, doing two different jobs.

• In Laura Aime’s case, the evidence yielded both a conventional autosomal STR profile and a Y-STR profile. The autosomal profile matched Bundy’s known profile directly. The known sample is a vial of blood drawn in Florida in 1978, after his arrest in the death of a twelve-year-old girl, and the full profile from it was entered into the FBI’s national database. Autosomal STRs are the standard forensic DNA markers, inherited from both parents, and they can individualize.
• In Melissa Smith’s case, the evidence yielded a Y-STR profile. The Utah State Crime Lab confirmed on July 14, 2026 that this Y-STR profile matched the Y-STR profile developed from the Aime evidence.

So the chain runs like this. Bundy’s known sample matched the Aime evidence directly, on markers that identify an individual. The Smith evidence matched the Aime evidence on a marker that identifies a paternal line. The link from Smith to Bundy is made by bridging the two cases together, not by a direct individual match in Smith’s case.

None of that is an inference. A forensic scientist manager at the Utah Bureau of Forensic Services described the marker as inherited paternally, father to son on down, and a Utah State Crime Lab analyst explained that a Y-STR profile traces DNA passed through the male lineage. Coverage that names the test includes Forensic Magazine, Newsweek, KUTV, TownLift and Cowboy State Daily, and the Associated Press told readers the DNA that identified Bundy came from the Y chromosome, which is only present in men. One local outlet went further and reported that the Y-STR result was not treated as a stand-alone identification, precisely because male relatives share the profile. That is the concession at the center of this post, made on the record, on the day, by the people who did the work.

Two further facts make the chain legible. The Aime evidence produced both profiles, so once the autosomal one matched Bundy there was a Y-STR profile sitting there available for comparison. And the sequence ran in that order: Florida reported the database hit in March 2026, the laboratory confirmed the Y-STR to Y-STR match on July 14, and Utah County presented the evidence to Summit County on October 7. Bundy was already in the frame before the Y-STR comparison was run, which is what makes that comparison a reasonable last step rather than the whole case.

For a man executed in 1989 who confessed to thirty murders, that is a perfectly sound way to reach a conclusion, and the prosecutor said expressly that she was relying on more than the DNA. The compression happened downstream. Several national outlets, CBS News, CNN and UPI among them, reported the closure without naming a test at all. And even where the test is named, “Y-STR profile links Bundy” is a different sentence from “Y-STR profile links the Smith evidence to the same paternal lineage as the Aime evidence.”

What a Y-STR Profile Is

The Y chromosome’s forensic markers do not recombine, so they pass down the paternal line as one block. They do not pass down unchanged, and it matters that they do not. Mutation rates run near one in a thousand per marker per generation, and because a profile combines twenty-three or twenty-seven markers, something like eight to thirteen percent of father-to-son transmissions change the profile somewhere. That slow drift is what lets Y-STR separate some relatives, and its slowness is what leaves many of them indistinguishable. Either way, a Y-STR profile is not a property of a person. It is a property of a male line.

A man shares his Y-STR haplotype with his father, his sons, his brothers, his paternal uncles, his paternal male cousins, and every other male on that branch of the family going back generations. According to the DNA Commission of the International Society of Forensic Genetics, paternal relatives have a high probability of sharing the same Y-STR profile, and the Commission reports a study in which testing male relatives across five generations with forty-seven Y-STR markers “cannot exclude them from being the donor of the trace evidence.”

The Commission gives figures, with its own hedge attached. Under certain assumptions, a relative twenty father-to-son steps away has roughly an 18 percent chance of carrying the same profile on a twenty-three locus kit, and about 5.5 percent on a twenty-seven marker kit. Twenty steps apart is a shared male ancestor around ten generations back, something on the order of ninth cousins, a kinship nobody alive would know about. And those are conditional probabilities about one specific relative, not the chance that some unrelated man matches, which for a modern kit is negligible.

Published estimates put the number of living men sharing a modern Y-STR profile in the tens: a median in the low teens to the mid twenties depending on the kit and the assumptions, with a ninety-fifth percentile that can run past a hundred. So a Y-STR match is powerful evidence. It takes the field from millions down to a few dozen. What it does not do is pick one man out of that few dozen, and the men in it are paternal relatives whose kinship is usually far too remote for anyone to recognize, trace or rule out by asking the family.

This is the paternal mirror image of something we have written about before. Mitochondrial DNA traces a maternal line and can exclude but cannot individualize. Y-STR is the paternal analogue, and both are reported as exclusions or non-exclusions rather than as identifications. They are not equal in power, and it is worth not overstating the symmetry. The mitochondrial genome mutates roughly ten times more slowly than a modern Y-STR kit, so mitochondrial matching groups are considerably larger and a Y-STR profile narrows the field much further. What the two share is the thing that matters here. Neither one points at one man.

The Math Does Not Work the Way Juries Think

Jurors who have heard a DNA number before have heard something like one in fifty trillion. Those figures come from autosomal STR testing, where the loci are on different chromosomes, assort independently, and can be multiplied together. That is the product rule, and it is what drives the numbers to astronomical sizes.

Y-STR loci sit on the same chromosome and are inherited as one block. They are not independent, so they cannot be multiplied. The ISFG Commission states the consequence directly: the Y-STR haplotype is “a single entity, for which the frequency needs to be estimated.”

That estimate comes from a reference database, usually the Y-Chromosome Haplotype Reference Database. The Commission ranks its two methods rather than treating them as equals. Its first choice is a model-based approach called Discrete Laplace. Augmented counting, which adds the query haplotype to both the observations and the database so the estimate can never be zero, is the fallback, and the Commission calls it easily defensible but highly conservative. Counting is preferred where the profile is partial, where it carries non-integer alleles, or where it runs to twenty-three or more markers, which Discrete Laplace cannot readily handle. Counting also has a floor: the Commission says counting estimators are bound from below by one over the database size. Which means a Y-STR statistic is limited by how many men are in the reference set, not by the power of the test.

The reference population matters, and the Commission says the choice should be driven by the origin of any possible suspect, not by the known defendant’s origin. It calls the latter “a common misconception,” and because the trace donor’s ancestry is unknown it treats the location of the crime scene as a leading criterion. A laboratory that picked its reference database to match the defendant has done the thing the Commission warns against. And the report wording matters. Where a laboratory reports a match qualitatively, without a frequency estimate, the Commission says the statement has to address male relatedness and the extent of haplotype sharing, and it separately says that information on the degree of relationship within the extended family has to be in the report where it is known.

One more thing the Commission recommends: Where a Y-STR profile matches, the Commission suggests typing rapidly mutating Y-STR markers on both the trace and the reference sample, specifically to try to exclude available close relatives. Those markers mutate fast enough that a single difference can separate men a conventional kit cannot. In the Commission’s own triage scheme it is the third priority, after autosomal testing and conventional Y-STR. If a laboratory stopped at a conventional Y-STR match and never ran a rapidly mutating panel, that is a question worth asking and a test worth requesting.

Notice what is missing from every account of the Utah announcement. No match probability. No haplotype frequency. No likelihood ratio. No database named. In a closed investigation of a dead man that omission costs nothing. In a courtroom it might be the whole argument.

Where You Will Actually Meet This in Texas

Y-STR testing is not an exotic cold-case tool. It is routine in sexual assault casework, and the reason is mechanical: when a sample contains a large amount of female DNA and a small amount of male DNA, autosomal testing can be swamped by the majority contributor. Y-STR primers target markers found only on the Y chromosome, so the female DNA does not amplify at all and the male profile is read free of it. It is often the only thing that produces a usable result.

That is a real benefit, and it has helped resolve enormous backlogs. It also means the lineage limitation shows up in live cases involving living defendants who have fathers, brothers, sons and cousins, sometimes in the same household.

If a Y-STR result appears in your case, these belong in a written request:

• The statistic and how it was derived. The haplotype frequency or likelihood ratio, the estimation method used, the database and its release version, and the reference population chosen, with the reason it was chosen.
• The number of loci tested and the kit used. A 17-locus result and a 27-locus result do not carry the same weight, and the sharing probabilities above change with the number.
• Whether autosomal testing was attempted first, and what it showed. The Commission’s triage puts autosomal STR analysis first, conventional Y-STR second where the autosomal result is inconclusive for the male component, and a rapidly mutating Y-STR panel third, expressly to exclude available close relatives.
• Whether any paternal male relative was ever tested or excluded. This is the question the science makes necessary and it is frequently never asked.
• The analyst’s report language. Whether it states that the result does not distinguish among male paternal relatives, and if it does not, why.
• The bench notes and electronic data, not the one-page conclusion.

Those requests run through Article 39.14, the Michael Morton Act, which in its current form applies only to the prosecution of an offense committed on or after January 1, 2014. For an older case the floor is Brady and nothing else, and that limit bites here, because destruction notices and post-conviction testing motions arrive in exactly the old cases. Where admissibility is the fight, Kelly v. State, 824 S.W.2d 568, 573 (Tex. Crim. App. 1992), is the Texas standard, and its three criteria are a valid underlying theory, a valid technique applying it, and proper application on the occasion in question. Do not reach for the third one here. A laboratory can execute Y-STR typing flawlessly and still produce a result that gets characterized far more strongly than the biology permits, and a correctly executed test offers no application defect to attack. The attack runs through relevance and Rule 403, where the controlling factor is the tendency of evidence to be given undue weight by a jury not equipped to evaluate its probative force, and through the first criterion recast as whether any valid theory supports individualizing a man from a shared paternal haplotype. Reserve the third criterion for a specific execution defect, such as a partial profile at too few loci or a contested mixture call. Texas has authority here: in Hernandez v. State, No. 09-16-00390-CR (Tex. App.-Beaumont Oct. 10, 2018) (mem. op.), a Y-STR challenge was litigated and resolved entirely under Rules 401 and 403 without citing Kelly at all, and the court described the result as consistent with any male in the defendant’s direct familial line. Note how the State got there. Relevance turned on the defendant conceding that no other male relative had access. Do not concede that. Article 38.35(d)(1) makes both the analysis and the expert testimony about it inadmissible where the laboratory was not accredited at the time, subject to an exception in subsection (e) that often cures a paperwork gap. And where a conviction already rests on testing that has since been reconsidered, two statutes are in play rather than one. Article 11.073 is the article for science that has changed or that now contradicts what the State relied on at trial, while Chapter 64 is the mechanism for getting a new test on evidence the State held at trial. Be clear about what that chapter does: it gets testing, not relief. It requires a sworn affidavit of supporting facts, a finding that identity was or is an issue in the case, and a showing by a preponderance that the movant would not have been convicted had exculpatory results been obtained, and relief then runs through a separate habeas application. We have written separately about how DNA interpretation software handles mixtures, which is the autosomal half of this same problem.

The Part That Cuts Both Ways

Salt Lake County Sheriff Rosie Rivera said this case demonstrates the importance of preserving evidence even when decades have passed, and that the work done today makes a difference decades from now. The detective who leads her cold case unit put the practice more plainly: in homicide cases they retain the evidence forever, and they keep it even in cases they have solved, in case something comes up down the road.

The scale behind that is real. In April a Utah Department of Public Safety spokesperson told the Salt Lake Tribune the state had 444 unsolved cases, including 258 homicides and 146 missing persons.

The preservation Rivera was praising is the same preservation that lets a wrongly convicted person get testing thirty years on. Every exoneration built on DNA depends on somebody having kept something. Nothing that was thrown away can be retested, and the person who suffers from a destroyed sample is at least as often the defendant as the State.

In Texas Article 38.43 requires any governmental or public entity, or an individual, charged with collecting, storing, preserving, analyzing or retrieving biological evidence, including law enforcement agencies, prosecutors, courts, public hospitals and crime laboratories, to retain it. But the retention period is two rules rather than one, and the forty-year figure is the narrower of them. Subsection (c)(1) sets the floor at not less than forty years, or until any applicable limitations period expires, only where an unapprehended actor is associated with the offense. Where the defendant was convicted, or placed on deferred adjudication or community supervision, and no actor is unapprehended, subsection (c)(2) ties the period to the disposition instead, which on a community supervision case can be a couple of years rather than forty. For the contents of a sexual assault examination kit, subsection (c-1) is the better rule: not less than forty years or until limitations expire, whichever period is longer, and it applies whether or not anyone has been apprehended or charged. Notice the conjunctions. Kit contents get the longer of the two periods. Other biological evidence does not.

The prosecutor, a clerk or any other officer holding the evidence may destroy it, and subsection (d) is thinner protection than it sounds. No court order is required and the statute does not condition destruction on the evidence ever having been tested. What there is, is notice to the defendant, to the defendant’s last attorney of record and to the convicting court, and then silence. Destruction is barred only if a written objection comes in from the defendant, the attorney of record or the court before the 91st day after the later of two dates: the day the custodian receives proof that the defendant got the notice, and the day the notice was mailed to the last attorney of record. Read that second leg again. It runs from the date of mailing, not from the date anybody reads it, so a notice sent to a stale address can start the clock in an office nobody is watching. No judge has to approve any of it. A closed file is exactly where a letter like that goes to die. Keep your address current with the courts, and if one reaches you, calendar it the day it arrives.

Being Fair About Utah

None of this is a criticism of the Utah investigation. Nobody is being prosecuted. The prosecutor did not claim the Y-STR result alone identified Bundy; she said expressly that she was relying on the DNA together with pattern, motive, means and opportunity. In a case with a confessed serial murderer, a second victim with a direct autosomal match, a geographic and temporal pattern and no living defendant, the conclusion is reasonable. Utah went further than that. Local reporting records that investigators did not treat the Y-STR result as a stand-alone identification, precisely because male relatives share the profile.

The problem is downstream. When a case like this is reported as a Y-STR profile linking a specific man to a specific murder, the next jury that hears the phrase “Y-STR match” brings that compression into the box with them. A juror who believes a Y-STR match names a man is a juror who has already decided something the science did not decide.

The Bottom Line

Two things are true at once, and a defense lawyer has to be able to recognize both.

Testing that did not exist in 1974 gave two families an answer fifty-two years later, and will keep doing that. The work is good and the results are real.

And a Y-STR profile identifies a paternal line rather than a person. It is powerful evidence, and saying so is the only way to be heard on the rest: it can take the field from millions down to a few dozen men. But it cannot distinguish a man from his brother, his father or his son. The few dozen are all his paternal relatives, and most of them are too distant for anyone to name. The statistics behind it are bounded by a database rather than by the product rule. And the right question when one appears in a live case is not whether it matches. It is who else it matches, and whether anybody tried to find out.

If you are facing a Texas case built on DNA evidence, our firm can talk with you about the specific facts.

Deandra Grant is the managing partner of Deandra Grant Law, a Texas DWI and criminal defense firm with offices across Dallas-Fort Worth and Central Texas. She holds a master’s degree in pharmaceutical science and a graduate certificate in forensic toxicology, and she completed DNA for Lawyers, a week-long workshop taught by Dr. Greg Hampikian at Boise State.

Further Reading

• Deseret News, The Salt Lake Tribune, CBS News and the BBC on the closure of the Melissa Ann Smith investigation (Oct. 27, 2026), and Forensic Magazine on the Y-STR testing.
• Roewer et al., DNA Commission of the International Society of Forensic Genetics: Recommendations on the interpretation of Y-STR results in forensic analysis (2020).
• SWGDAM, Interpretation Guidelines for Y-Chromosome STR Typing by Forensic DNA Laboratories (approved March 2, 2022, with a 2025 supplement). This replaces the January 2009 version published in the FBI’s Forensic Science Communications, which has now been superseded twice and should not be cited.
• Sexual Assault Kit Initiative, Y-STR Testing: Enhancing Sexual Assault Workflows.
• Kelly v. State, 824 S.W.2d 568 (Tex. Crim. App. 1992); Tex. Code Crim. Proc. art. 38.35 (accreditation and admissibility); art. 38.43 (preservation of biological evidence); art. 39.14 (Michael Morton Act); ch. 64 (post-conviction DNA testing).
• Andersen & Balding, “How convincing is a matching Y-chromosome profile?” PLoS Genetics 13(11):e1007028 (2017), with a March 2026 correction replacing the distribution of father-to-son steps between matching males. This is the source for the matching-group size, and its first author sat on the 2020 ISFG Commission.
• Bright, Andersen, Taylor, Kelly, Kruijver & Buckleton, “Relevant propositions for Y chromosome interpretation,” Journal of Forensic Sciences 70(1):271 (2025), which argues against framing the propositions around close paternal relatives. Anticipate it from a prosecution expert.
• UK Forensic Science Regulator, Guidance: Y-STR Profiling, FSR-GUI-0013 Issue 2 (March 2025), which adopts the ISFG approach and describes it as relevant but not mandatory.
• Hernandez v. State, No. 09-16-00390-CR (Tex. App.-Beaumont Oct. 10, 2018) (mem. op.) (Y-STR admissibility resolved under Rules 401 and 403); Lumsden v. State, No. 02-21-00012-CR (Tex. App.-Fort Worth Sept. 23, 2021) (mem. op.) (Chapter 64; a defense expert on the lineage limitation, and a reminder that the newer-technique showing has to be pleaded with specifics). Confirm publication designation before relying on either in a filing.
• Tex. Code Crim. Proc. art. 11.073 (new scientific evidence), which is the article for science that has changed, as distinct from Chapter 64, which is the mechanism for a new test.
• Related on this site: solving a cold case with 45-year-old hair, on mitochondrial DNA as the maternal lineage marker; and probabilistic DNA software in the courtroom.

This post is an informational synthesis for educational purposes and is not legal advice. Ted Bundy was executed in 1989 and no one is being prosecuted for the deaths described here; the conclusions stated are those of the Utah authorities who announced them. Descriptions of the testing come from news reporting, and no laboratory report has been published. Anyone facing charges should consult a licensed Texas attorney about the specific facts of their case.