Objective: To identify the factors associated to the number of medication changes (MC) in institutionalized elderly. Method: Prospective study conducted during four months. The number of MC was recorded weekly. We analyzed the factors associated with a high number of changes (four or more), calculating the Odds Ratio as a measure of association and the Fisher's exact test for statistical significance. Results: 997 MC were recorded in 234 elderly. 42.7% of the patients had four or more MC. The factors associated with a high number of MC were, chronic renal failure, immobility syndrome, loss of visual acuity, comorbidity, and polypharmacy. Conclusions: A high number of MC is associated with factors considered as risk indicators of drug-related problems. The number of MC in institutionalized patients may be an indicator to prioritize medication review.