Context: Women who rely on long-term hormonal contraception may neglec
t to use condoms, and thus increase their risk of contracting sexually
transmitted diseases, including AIDS. Methods: Data from a prospectiv
e, multisite study were collected to examine the probability of condom
use among 1,073 new users of either the contraceptive implant or inje
ctable; users were interviewed when they accepted their method and aga
in six months to one year later. Multivariate logistic regression anal
yses identified factors that significantly predicted the likelihood of
dual method use. Results: Condom use dropped markedly among women who
adopted long-term hormonal contraception. The proportion who always u
sed condoms in the previous three months fell from 21% at the time of
adoption to 11% at follow-up. Among women with one sexual partner, thi
s decrease was from 20% to 10%; however, among those with more than on
e partner, use increased from 25% to 31%. The factors significantly pr
edicting dual method use included previous condom use (odds ratio of 2
.5), receipt of AIDS-specific counseling (odds ratio of 1.6), the perc
eption of being at some risk of AIDS at baseline (odds ratio of 1.4) a
nd having had more than one sexual partner over the study period (odds
ratio of 5.4). In addition, injectable users, teenagers and black wom
en were more likely than other women to use condoms with their hormona
l method. Conclusions: Although condom use among all women declined ma
rkedly once they initiated long-term hormonal contraception, frequency
of condom use varied by subgroup and was associated with several fact
ors. Most importantly, women with more than one sexual partner and tho
se who received a message during counseling on the need to continue us
ing condoms were more likely than others to use condoms in conjunction
with the implant or injectable.