This contribution is a sequel to a previous article (Surgery 2000; 18(4): 100-4), in which I described the aim of evaluations of clinical effectiveness, the available study designs, threats to the validity of evaluations from selection and information bias and, in some circumstances, the difficulty of recruiting to randomized controlled trials (RCTs) involving one or more surgical treatments. This contribution should be read in conjunction with (and not be considered a replacement for) the previous article.