Background:
Young African American females are at highest risk for sexually transmitted infections (STI). Evidence suggests that the intersection of marijuana use and sexual behavior exacerbates adolescents' STI risk. The association between unprotected sex and marijuana use may be confounded by dispositional traits. Disposition may also increase the risk of choosing a high risk sex partner.

Objective:
To examine whether having a drug using sex partner is associated with condom use, independent of a girl's marijuana use and dispositional traits.

Method:
A prospective cohort of adolescent girls, aged 14 – 19 at baseline, were recruited from an adolescent health clinic or an STD clinic in Baltimore, MD and interviewed semi-annually for 3 years. Participants reported on their marijuana use, partner-specific condom use, and their main sex partner's drug use history at each semi-annual visit. Random effects logistic regression models were used to conduct subject-specific analyses on the repeated measures.

Result:
Girls varied in their marijuana use over the 3 year follow-up. Looking across a girl's main sex partners, we found that girls chose partners who were heterogeneous with respect to drug use. Girls were 3.74 times more likely to use marijuana (95% CI: 2.58, 5.41) and almost half as likely to use condoms (OR = 0.53, 95% CI: 0.37, 0.76) if their main sex partner was a drug user, holding girl's disposition constant. Partner drug use confounded the association seen between a girl's marijuana use and condom non-use.

Conclusion:
Longitudinal data suggest that individual disposition alone does not explain the association between marijuana use and condom non-use. Partner characteristics likely play a considerable role in the substance use and sexual risk behaviors of adolescent girls.

Implications:
Reducing a girl's marijuana use is not likely to impact her STI risk. Future research should investigate relationship qualities that contribute to marijuana use concordance.