The risk of acquiring HIV is particularly high among persons who use methamphetamine, which is often associated with unprotected sex and high numbers of sexual partners. This study examined the efficacy of a behavioral intervention emphasizing motivational interviewing and social cognitive theory for increasing safer sex behaviors in the context of ongoing methamphetamine use in a sample of HIV-negative, heterosexual methamphetamine users. Four hundred fifty-one participants from San Diego, California, were randomly assigned to receive one of three treatment conditions: (a) a safer sex behavioral intervention (Fast-Lane [FL]), (b) the FL intervention with boosters (FL + B), or (c) a time-equivalent diet-and-exercise attention-control (D&E) condition. Random effects regression analyses were used to evaluate change in safer sex behaviors over an 18-month period. Compared to those in the D&E condition, participants in the FL + B condition (p = .019) and FL condition (p = .020) significantly increased their engagement in protected sex acts over the active intervention phase. Also, compared to the D&E condition, those in the FL condition demonstrated a significant decrease in unprotected sex (p = .005) and an increase in percent protected sex (p = .001) during the active intervention. Finally, relative to D&E participants, FL participants demonstrated significant improvements in self-efficacy for negotiating safer sex (p = .011), and change in self-efficacy mediated the efficacy of the FL condition for increasing safer sex behaviors (p = .033). These results suggest that our behavioral intervention was successful in terms of reducing high-risk sexual behaviors in the context of ongoing methamphetamine use among HIV-negative heterosexuals. Reductions in high-risk sexual behavior were likely because of the impact of the intervention on participants' self-efficacy for negotiating safer sex.