A dog-scent lineup consists of matching a “scent” sample from a crime scene to a “scent” sample from a suspect by a dog. The practice has been used in several states, including Alaska, Florida, New York, and Texas. We know that dogs have an incredibly acute sense of smell, but the major problem has been with the handlers of these dogs, who have been proven to be fakes and charlatans.
The Innocence Project of Texas has published an excellent article about this practice in the state of Texas. While 71 pages in total, the actual article is only 14 pages – the rest being appended affidavits from experts. The link to the article is below:
The most infamous practitioner of this bogus science has been Deputy Keith Pickett of Fort Bend County, Texas. From 1994 to 2009 he traveled all over the state of Texas with his dogs, conducting dog-scent lineups. And he was always telling police and prosecutors exactly what they wanted to hear. The prosecutors loved him, and he was something of a “justice system rock star.” At one point, his status as an expert was even solidified by an appellate court decision.
The NY Times published an article about the wrongful imprisonment of Curvis Bickham in Texas. He was linked to a triple murder through bogus dog-scent lineup evidence provided by Keith Pickett and his dogs. Mr. Bickham was eventually released only because the real perpetrator confessed, and he had lost everything – his house, his cars, and his business. See the article here. And another NY Times article on the subject here. In 2009, the dog-scent convictions in Texas started being overturned. It’s about this time that Mr. Pickett “retired.” Nobody has records of exactly how may cases Keith Pickett and his dogs were involved in, but it’s believed to be as high as 2,000.
There is a similar situation in Florida with a dog handler named John Preston. See article here. To this day, nobody knows how many innocent people are still in prison as a result of John Preston’s fakery. Preston died in 2008 without ever having been charged for his fraud.
The most scientific approach to dog-scent lineups has been taken by the Dutch police, who have been establishing rigorous training & administration requirements and processes since the 1960′s. What they have found is that, even under the best of conditions and with the most rigorous processes, dog-scent lineup evidence is only 85% accurate. Dog-scent lineup evidence is not admissable in Dutch courts unless it is in conjuction with other evidence identifying the suspect. This article has more detail on the Dutch police practices with dog-scent lineups.