Regular long-term clinical follow-up is an important component of HIV care. To describe the reasons for follow-up loss among patients enrolled in the HIV/AIDS program of a university hospital. A nested case-control study was carried out on a retrospective cohort between January 1st, 2012 and July 31st, 2013. A group of 45 patients was selected; the incidence density rate of patients lost to follow-up was 17.7 per 100 patient/years. The following variables were significantly linked to follow-up loss in the bivariate analysis: Unemployment (p=0.000); alcohol consumption (p=0.004); number of years of evolution of the disease (p=0.032); gender (p=0.027), and mean age of 34 years (p=0.000). When logistic regression was adjusted for the probability of follow-up loss the significant variables were: Mean age of 34 years (p=0.019, 95% CI: 0.871-0.976); female (p=0.017, 95% CI: 1.903-31.83); alcohol consumption (p=0.028, 95% CI: 0.040-0.830), and unemployment (p=0.001, 95% CI: 4.696-464.692). HIV/AIDS programs need to establish follow-up systems and means to trace any losses in order to establish strategies to improve patient retention and, thus, their long-term quality of life.