Clinical trials are carefully designed experiments involving human subjects and aimed at improving patient care. They are of crucial importance in critical care medicine and their results can change clinical practice; however, methodological problems often exist and can compromise the validity of the study. Differences between single-centre and multicentre study design include variations in sample size, time of patient recruitment, costs, presence of bias, validity, centre and investigator characteristics. We reviewed methodological issues related to clinical studies, both single- and multi-centre, with special emphasis on the recent literature covering therapeutic interventions in critically ill patients. Multicentre collaboration can result in higher rates of patient enrolment than single-centre trials, thereby generating larger studies of shorter duration. Multicentre studies have rigorous study protocols to ensure uniform data collection among centres; however, heterogeneity in clinical practice among centres may be a major confounding factor in interpreting the results of these studies. The discrepancy between the results of single- and multi-centre studies may also be explained by variations in study protocols, inclusion and exclusion criteria, and adherence to study protocol. Factors related to timing of therapy, characteristics of the control group, and dosing of the study drugs may also be important. The results of multicentre studies should be interpreted with careful attention to their pros and cons. Copyright © 2008, Nederlandse Vereniging voor Intensive Care. All Rights Reserved.