OBJECTIVES: Adherence to antiretroviral therapy (ART) is a main concern in the control of HIV and is associated with treatment success or failure. The aim of this study was to determine adherence to ART using three methods and to identify factors related with non-adherence. METHODS: The study included 225 HIV patients under treatment at Regional Hospital in Concepción, Chile. Adherence was measured through two indirect methods: (1) the simplified medication adherence questionnaire (SMAQ), a survey including four questions with dichotomous answers, one question with a categorical answer, and one question with an open answer; and (2) a pharmacy dispensation register (PDR), which measures adherence through a ratio defined as number of days for which the supply of medication dispensed was prescribed divided by days between prescription fills; and through one direct method: (3) measuring viral load and CD4+ cell count. These methods were compared and a global weighted adherence measure was calculated for the 47 patients that received all three evaluations.  RESULTS: According to the SMAQ, 51% of patients were adherent; while 43% were adherent according to PDR, and 50% were adherent according to the direct method. Weighted adherence was 21%. Main causes of non-adherence were: delay in medication refill (43%); lack of information regarding ART (43%); forgetting doses (36%); and adverse reaction (35%). Patients were more adherent to treatment with the combination of two nucleoside reverse transcriptase inhibitors and one non-nucleoside reverse transcriptase inhibitor (33%) than with the same combination plus protease inhibitor (12%). CONCLUSIONS: Adherence to HIV treatment was below the Chilean average. Because the global weighted average was significantly less than the individual adherence percentages, each method alone is not sufficient to obtain a reliable determination of adherence. An evaluation of the impact of non-adherence behavior on the Chilean national HIV program is recommended.