BACKGROUND AND OBJECTIVE: To study the effect of clinical and demographic variables on mortality in patients with probable Alzheimer's disease treated with the cholinesterase inhibitor galantamine. PATIENTS AND METHOD: This retrospective cohort study reviewed 172 medical records, gathering information such as demographic and clinical variables, adverse events, number of withdrawals and duration of treatment with galantamine. RESULTS: Of 172 patients, 18.6% had adverse events. Galantamine was well tolerated in 15.4% of patients but they abandoned the treatment because of several reasons after a median duration of treatment of 13.3 months and an average dose of 15.0 mg/day. The overall rate of mortality was 12.5%, being, 19.0% for those who abandoned the treatment and 11.3% for those who did not. The univariate analysis showed that patients who died were older, had had more antipsychotic medications, had a higher total Blessed score and had suffered from more episodes of heart failure. The associated variables in the multivariate analysis using a binary logistic regression were mortality, sex, age, hypertension, heart failure, arrhythmia, antipsychotic treatment and greater cognitive impairment. CONCLUSIONS: The duration and the dose of treatment with galantamine were not associated with increased mortality. Related variables were an advanced age, male sex, cardiovascular diseases and antipsychotic treatment.