Objective Condom use behaviours are proximal to recent STI increases in the USA, yet it remains unclear whether the use of condoms has changed over time among unmarried, non-cohabiting young men who have sex with women (MSW) and how this variability is influenced by STI risk factors.

Methods To examine condom use over time among MSW aged 15–29, we used three cross-sectional surveys from the 2002, 2006–2010 and 2011–2017 National Survey of Family Growth. We estimated weighted percentages, adjusted prevalence ratios (APRs) and 95% confidence intervals (CI) to assess changes in condom use, stratified by whether MSW reported any STI risk factors in the past 12 months (ie, perceived partner non-monogamy, male-to-male sex, sex in exchange for money or drugs, sex partner who injects illicit drugs, or an HIV-positive sex partner).

Results We observed a divergence in trends in condom use at last sex between men aged 15 –29 with STI risk factors in the past 12 months and those without such history. We saw significant declines in condom use from 2002 to 2011–2017 among men with STI risk factors (APR=0.80, 95% CI 0.68 to 0.95), specifically among those aged 15–19 (APR=0.73, 95% CI 0.57 to 0.94) or non-Hispanic white (APR=0.71, 95% CI 0.54 to 0.93). In contrast, trends in condom use among men with no STI factors remained stable or increased. Across all time periods, the most prevalent STI risk factor reported was perception of a non-monogamous female partner (23.0%–26.9%). Post-hoc analyses examined whether condom use trends changed once this variable was removed from analyses, but no different patterns were observed.

Conclusions While STIs have been increasing, men aged 15–29 with STI risk factors reported a decline in condom use. Rising STI rates may be sensitive to behavioural shifts in condom use among young MSW with STI risk factors.