Introduction: Adverse drug reactions (ADRs) are an important cause of morbidity and mortality among hospitalized patients. Objectives: This study was designed to describe the frequency, severity, and causality of ADRs in internal medicine units at a third-level university hospital. Materials and methods: A descriptive study was performed at internal medicine units, by means of a structured format, review of clinical records, and interview of hospitalized patients. The Naranjo algorithm was applied to patient adverse events to define causality. Additionally, ADRs were classified according to the Rawlins and Thompson criteria. Results: One hundred patients (50 men and 50 women) were included in the study. Ninety nine (99) adverse events were found among the patients. The Naranjo algorithm was applied to adverse events, resulting in twenty nine (29) probable ADRs, twenty (20) possible ADRs and fifty (50) doubtful ADRs. Cardiovascular drugs and antibiotics were the most frequent therapeutic groups associated with ADRs. In addition, two preventable medication errors were identified. Conclusions: Frequency of ADRs was similar to the number reported in other studies in internal medicine units. It is necessary to systematize efforts of pharmacological surveillance in hospital wards, toward an opportune detection and prevention of ADRs.