At the present time, one of the main problems encountered by patients living with HIV is related to the notion of therapy adherence. In this context, quality of life appears to be an essential variable, because it seems correlated to adherent behavior. Mannheimer et al. (2005) have shown in a longitudinal study that adherent patients maintained a high level of quality of life over time. However, the long-term development of psychosocial variables related to quality of life in adherent patients at the time of their inclusion is not known. This study therefore aims, through longitudinal follow-up of a cohort of adherent patients, to describe the evolution over time of psychosocial factors related to quality of life. The sample was constituted from the HIV-infected population consulting in the respective wards of six cities in the Midi-Pyrénées region (France). Inclusion criteria were set according to objective and subjective criteria of adherence as recommended by the literature. The sample size was 133 individuals at T0 and 110 individuals at T1, which is compatible with chosen statistical requirements (two-tailed alpha risk 5%, study power 80%, SD 20). The results show a significant decrease over time in mental health and health perceptions in adherent patients. Moreover, at six months, results reveal, among patients who are no longer adherent, higher scores across several dimensions of stress, and a decrease in levels of mental health in comparison to adherent patients. It appears that living with the HIV for patients having an adherent behavior has repercussions over time, leading to a degree of frailty in mental health. Also, in adherent patients who become non-adherent over time, the stress variable appears to be essential.