Rwanda has achieved high enrollment into antiretroviral therapy (ART) programs but data on adherence after enrollment are not routinely collected. We used a mixed-methods approach (standardized questionnaires, pill counts, focus group discussions, and in-depth interviews) to determine levels of and barriers to ART adherence from the perspective of both patients and healthcare workers (HCW). Data were available from 213 patients throughout the first year on ART; 58 of them and 23 HCW participated in a qualitative sub-study. Self-reported adherence was high (96% of patients reporting more than 95% adherence), but adherence by pill count was significantly lower, especially in the first 3 months. In the standardized interviews, patients mostly reported that they "simply forgot" or "were away from home" as reasons for nonadherence. The qualitative research identified three interrelated constructs that appeared to negatively influence adherence: stigma, difficulty coming to terms with illness, and concealment of illness. Both standardized questionnaires and the qualitative research identified poverty, disruption to daily routines, factors related to regimen complexity and side effects, and service-related factors as barriers to adherence. We conclude that regular triangulation of different sources of adherence data is desirable to arrive at more realistic estimates. We propose that program monitoring and evaluation cycles incorporate more in-depth research to better understand concerns underlying reasons for nonadherence reported in routine monitoring.