Placebo controlled studies in patients suffering from exacerbation eg acute productive episode of schizophrenia and performed in the period between 1963-1993 are reviewed and analysed with respect to study designs; size of studies; improvement rate under placebo and drop-outs due to inefficacy under placebo. The aim of the analysis was to find out if the reported data permit some realistic estimates for a priori assumptions needed for proper planning of such studies, particularly in view of the numerous ethical and other difficulties which their performance encounters in the practice. Literature research revealed a rather limited number of rigorous, placebo-controlled, monotherapy studies (without intermittent or concomitant additional neuroleptics) in acute schizophrenia. Across comparison of findings from these studies was difficult due to differences in duration of treatment, assessment instruments and criteria of efficacy which illustrated a lack of methodological standards for studies in this indication. The improvement rate under placebo, if measured by Clinical Global Assessment (CGI) appeared, however, not to exceed 25%, whereas BPRS score reduction as efficacy criterion mostly provided higher response rates (up to 40%). Of interest however, is the finding that the response rate to conventional neuroleptics at the end of 4-6 weeks of treatment in some studies hardly exceeded 40%. The most sensitive measure of placebo effect seemed to be the drop-out rate due to inefficacy (up to 100%) and it is suggested to consider this measure and the survival analysis approach in designing future studies. The review demonstrated many unresolved methodological problems in testing antipsychotic drugs in acute schizophrenia and, particularly, the need of scientific evidence of validity and sensitivity of measures of antipsychotic efficacy. The findings reported up to now do not offer cues for rational estimates of the effect size differences between placebo and active drugs after short-term treatment in acute schizophrenia.