Therapeutic trials in multiple sclerosis, in part because of the marked intra- and inter-individual variability of its clinical course, are prone to serious flaws in both design and interpretation. As a consequence, it has been a common historical pattern that treatment regimens, which are enthusiastically recommended at one point in time, are later proven to be ineffective by more definitive studies. This review considers several recently published therapeutic trials in order to exemplify some of the difficulties that commonly arise in this area of clinical research.