As a pharmacy service giving pharmaceutical care at different levels of health care for elderly people, we needed a standardization procedure for recording and evaluating pharmacists' interventions. Our objective was to homogenize pharmacist interventions; to know physicians' acceptance of our recommendations, as well as the most prevalent drug related problems (DRP); and the impact of the pharmacists' interventions. To achieve this goal we conducted a one year prospective study at two levels of health care: 176 nursing homes (EAR) (8828 patients) and 2 long-term and subacute care hospitals (HSS) (268 beds). Pharmacists' interventions were recorded using the American Society of Health-System Pharmacists classification as the basis. Frequency of the different DRP and the level of response and acceptance on the part of physicians was determined. The Medication Appropriateness Index (MAI) was used to evaluate the impact of the interventions on the prescription quality. Patients' mean age was 84.2 (EAR) and 80.7 (HSS), and in both cases, polypharmacy 9 drugs was around 63-69%. There were 4073 interventions done in EAR and 2560 in HSS. Level of response: 44% (EAR), 79% (HSS); degree of acceptance of the recommendations: 84% (EAR), 72% (HSS). Most frequent DRP: inappropriate dose, length of therapy, omissions, and financial impact. Drugs for the nervous system are those with the most DRP. MAI values/medication improved from 4.4 to 2.7 (EAR) and 3.8 to 1.7 (HSS). A normalized way of managing pharmacists' interventions for different health care levels has been established. We are on the way to increasing collaborative work with physicians and we know which DRPs are most prevalent.