Methods used to construct a spatial-temporal sampling frame to recruit a probability sample of female street prostitutes in the 4,000 square mile area of Los Angeles County are described. Of 164 potential street areas named by informants or associated with marker establishments, 111 were judged to have sufficient prostitution activity for inclusion in the study. Completed interviews were obtained from 998 prostitutes, 638 of whom also gave blood samples, representing 85% of those from whom a blood sample was sought. Field results indicate that probability sampling techniques can be successfully used to study street prostitutes. The authors believe the sample areas identified account for most of the street prostitution in Los Angeles County. The response rate is within the range of 61 percent, assuming all women who refused screening or denied eligibility were simply refusing to participate, to 89 percent, if all such women were in fact ineligible.