The aim of the present study was to evaluate the objective and subjective efficacy as well as tolerability of olanzapine in acute treatment of schizophrenia spectrum disorders under naturalistic non-selective conditions. Inpatients with schizophrenia spectrum disorders, consecutively admitted over an 18-month period, treated with olanzapine, were included. Diagnoses were made according to ICD-10 criteria based on repeated clinical assessments. Efficacy and tolerability of olanzapine were assessed at baseline and at the end of inpatient acute treatment including Positive and Negative Symptom Scale (PANSS), Clinical Global Impression, subjective assessments, UKU and biological investigations. One hundred and twenty non-selected patients who met ICD-10 criteria for schizophrenia (73%), schizophreniform disorder (14%) or schizoaffective disorder (13%) were treated with olanzapine 15.3+/-5.2 mg/day. Baseline severity (PANSS total mean score 102.2) was higher compared to various admittance studies (PANSS total mean score 86-90). In 32% of patients (n=38), olanzapine treatment was discontinued, mainly because of inefficacy for positive (89%, n=34) and/or negative (95%, n=36) symptoms and/or because of adverse events (37%, n=14). Response rates as improvement in PANSS total score (after > or =3 weeks of treatment) of > or =20%, 30% or 40% were 68%, 55% and 35%, respectively. Response rates in post-hoc defined treatment resistant patients were not significantly different from non-refractory patients. Sedation (26%) was the most common side-effect, followed by weight gain (22%). With regards to subjective efficacy, 30% of the patients were not satisfied with the efficacy of olanzapine, while only 6% of the patients reported a not satisfying subjective tolerability. According to duration of olanzapine treatment, the results for patients, who remained in hospital, revealed a faster increase of weight compared to admittance studies (7 kg in 14 weeks versus 7 kg in 38 weeks). Olanzapine has been found to be effective and tolerable, also under naturalistic acute treatment conditions. Compared to previous double-blind admittance studies, patients had a higher severity of illness at entry and a lower > or =40% PANSS total score response rate. By contrast to previous results, mean dose of olanzapine was similar for multiple- and first-episode patients, and weight gain was more severe. The results underline the need of Phase IV studies for the assessment of clinical antipsychotic efficacy and tolerability.