To describe the relation between the level of cognitive impairment in Alzheimer's disease and the use of cholinesterase inhibitors (ChEIs) in neurology, geriatric and psychiatric units, and to establish the clinical profile of these patients. An epidemiological, multi-centre, cross-sectional study was conducted. Subjects included in the study were consecutive outpatients diagnosed with Alzheimer's disease, in accordance with the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders, and who had been treated with rivastigmine, donepezil or galantamine, either on its own or in association with memantine in the last six months. The recruitment period lasted three months. In a single visit, researchers determined the medication that was used, the dose, the mini-mental test, the overall clinical impression-overall improvement and the overall clinical impression-severity of the disease. A total of 1940 patients were selected from neurology, psychiatric and geriatric services all over the country. Possible differences in the habits of different specialists as regards prescribing were analysed, together with the relation between cognitive impairment and the type of medication employed. The mean age of the patients was 77 +/- 6.6 years, 62% of whom were females; the mean score on the mini-mental test was 17.4 +/- 5.5. The mini-mental score was similar in patients treated with rivastigmine (18.02 +/- 5.23), donepezil (17.08 +/- 5.54) or galantamine (17.34 +/- 5.38). In patients who were treated with memantine in association with a ChEI, the mini-mental score was significantly lower (11.44 +/- 5.68) (p < 0.0001). The doses of the different ChEIs used by the specialists were similar. A higher percentage of patients had maximum doses of donepezil (81%) than in the cases of rivastigmine (43%) and galantamine (67%). The different specialists involved (neurologists, geriatricians and psychiatrists) displayed similar habits regarding the utilisation of ChEIs to treat Alzheimer's disease. There was no relation between the degree of impairment and the drug chosen, except in the case of memantine.