OBJECTIVES: To estimate changes in measures of hospitalisation use in a naturalistic clinical setting in Finland following switch to Risperdal Consta, a long-acting atypical antipsychotic. METHODS: Data were collected retrospectively from patient charts at 10 geographically and functionally diverse sites in Finland. Patients were at least 18 years old, diagnosed with schizophrenia or schizoaffective disorder, and initiated treatment with Risperdal Consta between January 1, 2004 and June 30, 2005. The study employed a mirror-image design. Two analyses were conducted: 1) Modified intention-to-treat (ITT) analysis based on the complete observational dataset, and 2) Matched per-protocol analysis, with duration of pre-switch patient data matched to the duration of Risperdal Consta treatment follow-up. The main outcome measures are the change in mean rate of hospitalisations per year between the pre- and post-switch periods and change in mean annual number of days in hospital. RESULTS: A total of 177 patients fulfilled the inclusion criteria (mean age 47 years, 52% female, average duration of schizophrenia 15 years). The most common reason for switching to Consta was non-compliance on other medications (63%). Consta treatment continuation rates of at least 6, 12, 18 and 24 months were 76.6%, 70.9%, 68.0% and 66.2%, respectively. The ITT analysis shows that switch to Consta was associated with significant reductions in number of hospitalisations per year (from 0.93 to 0.74) and in mean annual days in hospital (from 62.9 to 38.0 days per patient-year of treatment), corresponding to an estimated annual cost saving of €11,948 per patient. The matched per-protocol analysis shows larger changes for all endpoints compared with the ITT analyses. All changes were statistically significant at the 95% confidence level. CONCLUSIONS: This study of 177 patients with schizophrenia switching to long-acting Risperdal Consta indicates a high treatment continuation rate and sizeable reductions in inpatient resource use.