This study examined the effect of time lag on the validity of retrospective self-reports of sexual behavior. Seventy-five heterosexual students (44 women, 31 men) made daily recordings of sexual behavior, condom use, and alcohol or substance use for 1 month. Ability of respondents to recall sexual behavior recorded during this period was assessed at 1, 2, and 3 months after diary completion using recall interviews (25 interviewed at each interval). For vaginal intercourse, total recall error was significantly greater at 3 months than at 1 month post-diary. For all other variables assessed, the 2- and 3-month time intervals did not produce significant increase in total recall error. Higher frequency of vaginal intercourse, orgasm, and alcohol use prior to sexual activity were associated with total recall error for some but not all behaviors and outcomes. The results provide a partial validation of the diary-interview recall model as a method for studying recall error.