BACKGROUND: Most new HIV infections are attributed to male-to-male sexual contact in the United States. However, only two-thirds of sexual minority men living with HIV achieve an undetectable viral load (UVL). We tested a web-based antiretroviral therapy adherence intervention called Thrive with Me (TWM) with core features that included medication self-monitoring and feedback, HIV and antiretroviral therapy information, and a peer-to-peer exchange. OBJECTIVE: We assessed the efficacy of TWM on HIV UVL among adult (aged ≥18 years) sexual minority men. Moreover, we assessed the impact of overall engagement and engagement with specific intervention features on HIV UVL. METHODS: In total, 401 sexual minority men (mean age 39.1, SD 10.8 y; 230/384, 59.9% African American) in New York City were recruited between October 2016 and December 2019 and randomized to receive TWM (intervention) or a weekly email newsletter (control) for 5 months. Computerized assessments occurred at baseline and months 5, 11, and 17. The primary outcome was a dichotomous measure of HIV UVL (≤20 copies/μL). Generalized estimating equations with robust SEs were used to assess the effect of the TWM intervention on HIV UVL over the follow-up period in an unadjusted model and a model adjusted for baseline differences and then stratified by baseline recent drug use urinalysis. In secondary analyses, generalized linear models were used to estimate risk differences in the association of overall engagement with TWM (the sum of the number of days participants accessed ≥1 screen of the TWM intervention out of a possible 150 days) and engagement with specific TWM components on HIV UVL throughout the 17-month intervention period. RESULTS: Participant retention was 88.5% (355/401; month 5), 81.8% (328/401; month 11), and 80.3% (322/401; month 17). No consistent differences in HIV UVL were found between those randomized to receive TWM or the control at the 5- (difference-in-differences [DD]=-7.8, 95% CI -21.1 to 5.5), 11- (DD=-13.9, 95% CI -27.7 to 0.04), or 17-month (DD=-8.2, 95% CI -22.0 to 5.7) time points, or when stratified by baseline recent drug use. However, those TWM-assigned participants with high overall levels of engagement (in the upper 25th percentile) were more likely to have an HIV UVL at the end of the 5-month active intervention period compared to those with low engagement (below the 75th percentile; risk difference=17.8, 95% CI 2.5-33.0) or no engagement (risk difference=19.4, 95% CI 3.3-35.5) in the intervention. Moreover, high engagement with the peer-to-peer exchange was associated with HIV UVL over time in unadjusted models. CONCLUSIONS: TWM did not have overall impacts on HIV UVL; however, it may assist some sexual minority men who are highly engaged with this web-based intervention in achieving HIV viral suppression. TRIAL REGISTRATION: ClinicalTrials.gov NCT02704208; https://clinicaltrials.gov/study/NCT02704208.