The methodological quality of 50 clinical trial protocols submitted to our hospital has been assessed by means of a check-list. The most frequent methodological deficiencies found were related to statistical analysis, selection criteria, sample size, incorrect use of placebo, homogeneity of the groups, concomitant medication, randomisation plan, monitoring of adverse events and study design. Lack of insurance for the patients and inadequacies in the investigators' brochure and case report forms were observed in a significant number of cases. The results suggest the importance of a multidisciplinary team in the elaboration of clinical trial protocols to prevent methodological errors.