The study population included HAPPENS clients seen during a four and a half year data collection period from 1994 to 1998. Because we were interested in predictors of longer-term retention in care, those clients who had no more than the two standard HIV counseling and testing visits were excluded from our analyses.