Background Cognitive behavioral therapy (CBT) is frequently used to treat depressive symptoms in people living with HIV. We developed an internet-based cognitive behavioral intervention for people with HIV and depressive symptoms, which was based on an effective self-help booklet. The Web-based intervention was previously found to be effective. Objective The objective of this study was to investigate potential mediators of the Web-based intervention. Methods This study was part of a randomized controlled trial, in which the intervention was compared with an attention-only waiting list control condition. Participants were 188 (97 in intervention group and 91 in control group) people with HIV and mild to moderate depressive symptoms recruited in HIV treatment centers in the Netherlands. A total of 22 participants (22/188, 11.7%) in the study were female and 166 (166/188, 88.3%) were male. The average age of the participants was 46.30 years (SD 10.63). The intervention comprised Web-based self-help CBT for 8 weeks, 1 to 2 hours a week, including minimal telephone support from a coach. The participants received Web-based questionnaires at pretest, 3 times during the intervention/or waiting period, and post intervention. The outcome was depressive symptoms. Factors tested as potential mediators were changes in behavioral activation, relaxation, the cognitive coping strategies catastrophizing and positive refocusing, goal re-engagement, and coping self-efficacy. Results Using multilevel structural equation modeling, changes in behavioral activation (P=.006) and goal re-engagement (P=.009) were found to be significant mediators of the intervention effect. The mediation effect seemed to occur between weeks 3 and 5 for behavioral activation and weeks 1 and 3 for goal re-engagement. Using (bivariate) autoregressive latent trajectory analysis, we found a return effect (from the dependent variable to the mediator) for goal re-engagement but not for behavioral activation, which suggested that the mediation effect of changes in behavioral activation was stronger than that in goal re-engagement. Conclusions The results suggest that changes in behavioral activation and goal re-engagement may mediate the effect of the Web-based intervention for people with HIV and depressive symptoms. The results may lead to possible mechanisms of change of the intervention and improvement of therapy outcomes. Clinical Trial Netherlands Trial Register NTR5407; https://www.trialregister.nl/trial/5298