DNA from a crime scene cannot always be matched to a known individual in a database, but forensic investigators could identify close relatives.
DNA techniques are getting ever more sophisticated
To date, the controversial method has been tried in a small number of cases, sometimes to dramatic effect.
But the technique has the potential to be used much more widely, three US experts argue in the journal Science.
As DNA databases expand, policy makers need to consider the method's ethical and legal dimensions, they warn.
All humans have some genetic similarity, but close relatives have particular similarity because of their shared ancestry.
"Normally one would look for a perfect match between crime scene [DNA] evidence and a known offender in a database," co-author Frederick Bieber, of Harvard Medical School in Boston, US, told the Science In Action programme on the BBC World Service.
"Very close but not perfect matches might indicate, with some reliability, that crime scene evidence was derived from somebody very closely related to somebody in the databank.
"The offender may not be in the databank, but he may be the father or the son or the brother of someone who is."
Dr Bieber along with colleagues Charles Brenner and David Lazer carried out mathematical simulations to assess the potential of "kinship analysis" for identifying promising leads in forensic investigations.
"We simulated real families using the genetic frequencies of the markers used in forensic investigations throughout the world," Dr Bieber said.
"We asked the question mathematically, 'how often could you find a close relative - namely a parent, child or sibling - of the perpetrator of a crime?' Our data show that this would identify a close relative with a very high frequency."
Familial searching was crucial to police solving the 1988 murder of 16-year-old Lynette White in Cardiff, UK.
A search of the National DNA Database for a rare gene variant found in a specimen recovered from the crime scene identified a 14-year-old boy with a similar genetic profile.
This led police to his paternal uncle, Jeffrey Gafoor, the murderer. Gafoor received a life sentence in 2003 for the killing.
However, some researchers have expressed worries over how familial searching might be applied on a larger scale.
"My concern is that this evidence, if it ever comes to that, would be totally mishandled; and, in all likelihood, its value would be overstated by our forensic labs," commented Dr Laurence D Mueller, an evolutionary biologist at the University of California, Irvine.
"When you look through a database, one of the things people are concerned about is that you are comparing the genetic profile of thousands of people to the profile in an evidence sample.
"Just as it is more likely that you will win the lottery if you buy many tickets, people have been concerned that you're more likely to coincidentally find a match [with an innocent person] when you search a database."
The Science magazine authors themselves note that evidence of a DNA match can be misleading and does not necessarily prove guilt.
In addition, they say, this novel technique potentially amplifies existing disparities in the criminal justice system where arrests and convictions differ widely based on race, ethnicity, geographic location and social class.