DNA technologies have radically reshaped the role of forensics in police work. Even small amounts of blood, saliva, or other biological materials left at a crime scene can now lead to the identification or elimination of a suspect. Genetic evidence has been used both to convict perpetrators and to exonerate people who were wrongfully convicted on less reliable evidence, including scores of people on death row.

DNA typing is typically quite accurate when used to tell whether an unknown sample matches another sample that has already been identified. This is not to say that this process is without problems; simple human error, sample contamination, and misinterpretations have been known to skew results.

The development of forensic DNA databases--in which hundreds of thousands of profiles are stored with the intention of catching recidivists--has given rise to new sets of problems such as miscalculations of the statistical probability that an unknown sample coincidentally matches a stored profile. In some cases, what are touted as rare "one-in-a-million" odds of being a coincidental match are actually significantly more likely once other relevant factors (such as database size) are taken into consideration. Such information has, on occasion, not been revealed to juries.

Nevertheless, the compilation of DNA databases has been increasing dramatically. In many jurisdictions, both in the US and abroad (especially in the UK), they now include people who may have been arrested for but never convicted of a crime. This raises privacy issues in addition to issues of racial discrimination since minorities have disproportionately higher contact with police and are therefore overrepresented in these databases.