Adherence to medications is critical to optimizing HIV care and is a major challenge in youth. The utility of directly observed therapy (DOT) to improve adherence in youth with HIV remains undefined and prompted this pilot study. Four U.S. sites were selected for this 24-week cooperative group study to assess feasibility and to identify the logistics of providing DOT to HIV-infected youth with demonstrated adherence problems. Once-a-day DOT was provided by DOT facilitators at the participant's choice of a community-based location and DOT tapered over 12 weeks to self-administered therapy based on ongoing adherence assessments. Twenty participants, median age 21 years and median CD4 227 cells/microl, were enrolled. Participants chose their homes for 82% of DOT visits. Compliance with recommended DOT visits was (median) 91%, 91%, and 83% at weeks 4, 8, and 12, respectively. Six participants completed >90% of the study-specified DOT visits and successfully progressed to self-administered therapy (DOT success); only half sustained >90% medication adherence 12 weeks after discontinuing DOT. Participants considered DOT successes were more likely to have higher baseline depression scores (p = 0.046). Via exit surveys participants reported that meeting with the facilitator was easy, DOT increased their motivation to take medications, they felt sad when DOT ended, and 100% would recommend DOT to a friend. In conclusion, this study shows that while community-based DOT is safe, feasible, and as per participant feedback, acceptable to youth, DOT is not for all and the benefits appear short-lived. Depressed youth appear to be one subgroup that would benefit from this intervention. Study findings should help inform the design of larger community-based DOT intervention studies in youth.