OBJECTIVES:
We examined parameters of sexual partnerships, including respondents' participation in concurrency, belief that their partner had concurrent partnerships (partners' concurrency), and partnership intervals, among the 2099 women in HIV Prevention Trials Network 064, a study of women at high risk for HIV infection, in 10 U.S. communities.

METHODS:
We analyzed baseline survey responses about partnership dates to determine prevalence of participants' and partners' concurrency, intervals between partnerships, knowledge of whether recent partners had undergone HIV testing, and intercourse frequency during the preceding 6 months.

RESULTS:
Prevalence of participants' and partners' concurrency was 40% and 36%, respectively; 24% respondents had both concurrent partnerships and nonmonogamous partners. Among women with >1 partner and no concurrent partnerships themselves, the median gap between partners was 1 month. Multiple episodes of unprotected vaginal intercourse with ≥2 of their most recent partners was reported by 60% of women who had both concurrent partnerships and nonmonogamous partners, 50% with only concurrent partners and no partners' concurrency, and 33% with only partners' concurrency versus 14% of women with neither type of concurrency (P < 0.0001). Women who had any involvement with concurrency were also more likely than women with no concurrency involvement to report lack of awareness of whether recent partners had undergone HIV testing (participants' concurrency 41%, partners' concurrency 40%, both participants' and partners' concurrency 48%, neither 17%; P < 0.0001).

CONCLUSIONS:
These network patterns and short gaps between partnerships may create substantial opportunities for HIV transmission in this sample of women at high risk for HIV infection.