The objective of this study was to estimate the prevalence of antipsychotic use and investigate their association with behavioural and psychological symptoms of dementia (BPSD) and other clinical predictors. Patients with dementia, aged 65 and above and resident in 35 Alzheimer special care units were sequentially enrolled into a 18-month prospective observational study. Data on sociodemographic, cognitive, functional, behavioural and clinical characteristics and drug exposure were collected at baseline and at 6-month intervals up to 18 months. The prevalence of antipsychotic use and the association with BPSD and clinical predictors were analysed. Of the 349 patients with dementia enrolled in the study, 209 (60%) were taking at least one antipsychotic. Risperidone and promazine were the most frequently prescribed antipsychotic; 40.7% simultaneously received a benzodiazepine, 20% an antidepressant. More than 50% were still taking antipsychotics at 18 months of follow-up. No associations were found between antipsychotic use and level of cognitive impairment, basal activity of daily living disability and comorbidity. Multivariate analysis showed that the use of antipsychotics was highest in patients in the highest quartiles of Neuropsychiatric Inventory Scale score (III quartile, odds ratio: 1.63; 95% confidence interval: 1.19-2.23; IV quartile, odds ratio: 2.27; 95% confidence interval: 1.61-3.26). This study found high rate of use of antipsychotics in patients with dementia resident in Alzheimer special care units, frequent associations with other psychotropic medications and a strong correlation with BPSD.