Evidence indicates that regular assessment of antiretroviral therapy (ART) adherence is necessary to promote adherence and achieve viral suppression. Self-reported adherence using a visual analog scale (VAS) has been used extensively to measure ART adherence. However, less is known about the accuracy of the VAS for measuring ART adherence among opioid-dependent people living with HIV. In this study, we aimed to evaluate the accuracy of the VAS in measuring ART adherence in opioid-dependent people living with HIV who are enrolled in a methadone maintenance program (MMP). This study was conducted within a larger randomized controlled trial among opioid-dependent people living with HIV (N = 109) who were enrolled in an inner-city MMP. Self-reported VAS ART adherence data were compared to pharmacy refill data, which is a more objective measure. Self-reported VAS ratings and pharmacy refill behavior were significantly correlated at the 6-month post-intervention mark (r = 0.349, p = 0.006) but not at any other time point (i.e., pre-intervention, post-intervention, 3-months, and 9-months post-intervention). Based on the pharmacy refill data, participants did not consistently adhere to their ART medication but self-reported high adherence on the VAS. The self-report VAS may not be an accurate measure for assessing ART adherence among opioid-dependent persons living with HIV and enrolled in a MMP. Based on our findings, it may be more accurate to use pharmacy refill information or other objective measures to quantify ART adherence over extended periods for this patient population.