The present study is a critical review of 100 papers concerning clinical trials in Multiple Sclerosis (MS), published between 1970-1990. We analyzed 20 validity criteria (internal 12, and external 8) and codified each item as "adequate" or "not adequate", so that the validity of each report was quantified. In addition, we used a statistical method of automatic classification (cluster analysis) to subdivide all the papers in 4 groups with growing validity. Only 52 reports had a fairly good or good validity, therefore more efforts are suggested to correctly plan clinical trials in MS and lead to reliable results.