Dogs vs. DNA

K-9 testimony debunked

In dozens of cases in the 1980s, Brevard County, Florida, police officer John Preston testified that his police dogs could not only track incriminating scents over long distances and through water, but pick them up months and even years after a crime took place. Now three men convicted based on evidence from Preston's dogs have been exonerated, and a criminal justice advocacy group claims to have found a fourth.

A team of Texas hounds has a similar record. Fort Bend County police dogs, with the names of James Bond, Quincy, and Clue, recently picked two men out of a "scent lineup" under the supervision of Sheriff 's Deputy Keith Pikett. Both men were later cleared, one after serving three months in jail. In June they filed suit against Pikett.

These aren't isolated incidents. In a 1990 Hastings Law Journal article on police dog identifications, University of Pittsburgh law professor Andrew Taslitz concluded that police, prosecutors, and judges across the country had been relying on "unscientific myth." In 2006 another law professor, Richard Myers of the University of North Carolina, conducted a statistical analysis of police dog accuracy tests, and concluded that they weren't reliable enough to produce probable cause for search warrants, much less serve as the cornerstone of a conviction.

That doesn't mean dogs are going off the beat. In May, the South Carolina Supreme Court affirmed the evidentiary value of police hounds in criminal cases.