Television detectives relish pulling a single hair from a crime scene, like the proverbial needle from a haystack. "If only the perp hadn't shed this one strand," they say, "he would've gotten away with it." In real life, though, a strand of hair is not the smoking gun CSI may have you believe. While hair evidence may be valuable in narrowing a field of suspects, matching it to one particular individual is not at all a straightforward task. Forensic scientists use a variety of indicators--color, length, thickness, whether hair has been dyed or curled--to assign a hair to an individual, but there is no threshold for how many features must be the same in order to declare a match. Even when the examiners determine a match under a microscope, the potential for error is high. A 2002 FBI study discovered microscopic matches to be wrong 12.5% of the time when subsequently checked with mtDNA analysis.