Bayes' theorem deals with the ratio of two probabilities: the conditional probability of the evidence calculated under the assumption that the suspect did commit the crime divided by the conditional probability of the evidence calculated under the assumption that the suspect did not commit the crime. In order to understand the probability of the evidence under the latter assumption, a survey of people unassociated with the crime is warranted. In many instances, such survey data are already available. The dilemma arises with the choice of a particular population. A survey should be representative of all possible offenders using background information such as race, nature, or lifestyle of the suspect only if relevant to the case. In cases where there is a transfer from the offender to the scene, the race, nature, or lifestyle of the suspect do not influence thinking on which population to choose because the suspect details are irrelevant. However, data developed from previous cases should not be dismissed and may prove useful in cases where the transfer was from the scene to the suspect. In these cases, the nature and lifestyle of the suspect condition the thinking about selection of population type. 6 references