Crime scene investigators may soon have a new tool—one that could let them determine a suspect or victim’s age from a bloodstain.

Researchers in the Netherlands have developed a test that quantifies the amount of a specific molecule in blood to approximate a person’s age to within nine years, plus or minus. That’s obviously not specific enough to pick a perpetrator out of a lineup. But Victor Weedn, the pathologist for Maryland’s Office of the Medical Examiner, say it’s close enough to help identify missing persons and victims in large-scale disasters.

“For identification in mass disasters, it would clearly be useful to have age determinations,” says Weedn. “For criminalistic testing, it would help when you have two suspects, an older and a younger suspect—think the Beltway sniper—but you do not have direct references on them yet.”

Investigators can already determine a person’s age by analyzing bones and teeth. But most often, bloodstains are the only evidence at a crime scene. And while scientists can analyze DNA in blood to identify characteristics like eye color, skin tone, and gender, age is a trickier issue.

Manfred Kayser, head of the forensic molecular biology department at the Erasmus MC University Medical Center Rotterdam in the Netherlands, says there is no DNA marker for age. “You have a problem because normally your DNA does not change with age,” says Kayser. “What does change is the activity of certain genes over a lifetime, but that is at the RNA level, and crime labs aren’t ready for that. So we really had to look for another approach.”

Kayser and his colleagues dug through the scientific literature, looking for evidence of molecular changes in blood associated with aging. They came upon an age-related phenomenon in the thymus, an organ that produces a variety of immune cells called T-lymphocytes, which circulate in the blood to help the body fight off infections and other foreign invaders. Tiny molecules called signal joint TCR excision circles (sjTREC) help determine the type of T-cells produced in the thymus. From birth, the thymus immediately starts to shrink, producing fewer T-cells and sjTREC molecules with age.

Kayser hypothesized that sjTREC in blood could be an effective marker for age. He and his team developed a way to quantify the amount of sjTREC in blood, and analyzed 200 blood samples from people ranging from a few weeks old to 80 years old. They could reliably predict a person’s age, plus or minus nine years. Kayser took a second sample from volunteers and, after storing the samples for a year and a half, ran them through the same test, with similar results.

“The results were basically identical,” says Kayser. “We wanted to prove that if a blood sample lies around, that this time doesn’t influence the age of the person from the sample.”

The results are published this week in the journal Current Biology.

“The science behind this is innovative, and the test itself is a good addition to the forensic arsenal,” says George Sensabaugh, professor of biomedical and forensic sciences at the University of California at Berkeley. “However, the test likely would be useful only in a limited number of cases.”

Kayser and his team are looking for other signs of aging, both molecularly and genetically, to more specifically identify a person’s age.

“Estimates of age, of even 20-year increments, are valuable,” says Weedn. “With more technology, we may get the 20-year window to narrow. This is a start.”