Objective Over 30% of HIV-infected adults and adolescents in the United States are female, yet women make up less than 20% of participants in most antiretroviral clinical trials. We sought to understand how policies regarding fertility regulation might have influenced women's participation in Adult AIDS Clinical Trials Group (AACTG) antiretroviral trials. Methods We reviewed AACTG and the National Institute of Allergy and Infectious Diseases Division of AIDS (DAIDS) policies relating to the eligibility of women of reproductive potential to participate in antiretroviral clinical trials from 1996 to the present. We focused specifically on policies related to the documentation of women's reproductive potential, contraceptive requirements stratified by the FDA Use-in-Pregnancy categories, the management of incident pregnancies for female study subjects, and the impact of previous limited antiretroviral use in pregnancy and eligibility for antiretroviral-naive clinical trials. Results Before 1998, most AACTG protocols specified that ‘adequate birth control’ should be used during clinical trials and for 3 months afterwards, but there was no standard definition of adequate birth control. Most protocols implied that it was the woman's responsibility to prevent pregnancy. In 1998, a focus group on pregnancy issues and antiretroviral agents recommended that adequate birth control should be practised by both heterosexually active women and men on study. Adequate birth control was defined as either abstinence, hormonal plus barrier methods, or two barrier methods. In 2000, women of reproductive potential were explicitly defined as all adult and adolescent women except those who have undergone surgical sterilization, hysterectomy, bilateral oophorectomy or menopause, or whose male partners have documented azoospermia from a vasectomy or other condition. Source documentation of sterilization, menopause or azoospermia was required to confirm a lack of reproductive potential for study subjects. Women of reproductive potential were required to use at least one non-hormonal contraceptive method for trials involving FDA category A, B and most category C drugs, and were required to use two non-hormonal contraceptive methods for some category C drugs (efavirenz) and all category D and X drugs. Hormonal contraceptive methods were thought to be possibly ineffective when prescribed with some antiretroviral agents, and so could not fulfill the contraceptive requirement. IUD were discouraged because of the potential increased risk of pelvic inflammatory disease. In 2002, the source documentation requirement for women not of reproductive potential was relaxed for trials of category A, B and C drugs with the exception of efavirenz; however, the contraceptive requirements were not changed. Most antiretroviral trials require that female study subjects who become pregnant discontinue study medications unless there are adequate safety data in pregnancy. Most clinical trials for antiretroviral-naive study subjects exclude women with limited previous exposure to zidovudine or nevirapine during a pregnancy. Conclusion The vast majority of HIV-infected women in the United States are of child-bearing age, and so fertility regulation policies may have a substantial impact on women's eligibility for and participation in antiretroviral clinical trials. The burden of pregnancy prevention disproportionately affects women, including those who are not heterosexually active but are still thought to be ‘of reproductive potential’. We need better and more complete data regarding the safety, pharmacokinetics and efficacy of newer antiretroviral agents in pregnancy and in conjunction with hormonal contraceptives. In addition, antiretroviral-naive clinical trials should consider whether the exclusion of women with previous limited antiretroviral exposure in pregnancy is truly justified.