Outcome measurement tools were almost exclusively used, so far, in research studies, while in routine clinical practice gross descriptive evaluation was generally provided. Discrepancies among studies and among stakeholders as to the effectiveness of one or the other treatment, study designs based on strict inclusion and exclusion criteria unlikely to generalize to clinical practice, the choice of outcome measures not always clinically meaningful, difficulties in the interpretation of clinical trials results and the relatively short trial duration have contributed to clinicians’ skeptical attitude toward clinical trials findings. Researchers designing future clinical trials are confronted with the need of providing clinically meaningful outcome measures, without undermining the statistical power of the study. However, integrative measures, such as "time to discontinuation for any reason", thought to convey information about efficacy, tolerability and safety, can be influenced by several nonspecific variables likely to vary from site to site and from doctor to doctor. Simple and global measurements, such as CGI-SCH, while preferred by many clinicians, might not capture the non unitary nature of the schizophrenia outcome constructs. Remission criteria are still limited to symptoms and clinicians would certainly like to see criteria for functional remission implemented as well. Clinical trials based on simple designs and on more comprehensive outcome measures will probably contribute to reduce the gap between research findings and clinical practice.