The lack of a medical cure or preventive vaccine for HIV calls for int
erventions that reduce the occurrence of the behaviors known to be ass
ociated with infection. The development of effective interventions req
uires rigorous evaluations that, to date, have been noticeably lacking
in the field. In particular, assessments of HIV prevention efforts ha
ve considered interventions as ''black boxes,'' with little attention
to the actual services delivered. In this paper, issues related to eva
luating outreach interventions are presented and an evaluation strateg
y to measure the delivery of outreach services, using both quantitativ
e and qualitative techniques, is recommended.