A check list is described which helps the systematic assessment of reports of therapeutic trials, particularly the aspects that need to be considered in assessing their validity. The check list was used to examine 141 reports of therapeutic trials published in four British non-specialist journals in the first six months of 1966 and of 1969. Of these reports 51% were found to be acceptable and a further 16% probably acceptable; 33% were considered unacceptable because they lacked one or more of the features required in a valid report. The check list has been found useful in assessing claims made for drugs and other therapeutic measures on the basis of published reports.