In order to ensure sufficient disease activity, patients with relapsing remitting (RR) multiple sclerosis (MS) are often included in randomized placebo-controlled trials, only if they have a high baseline activity. These patients, whose evolution is unusual in the pre-study period, will tend to show a more usual behavior when followed up over a period of time. This phenomenon is known as regression to the mean. Regression to the mean should be taken into account in correctly interpreting long-term studies of cohorts treated without a placebo control group, which use the baseline period as control. The aim of this study was to evaluate the relevance of this phenomenon in a non-treated cohort of RRMS patients, selected with similar criteria to those used in randomized placebo-controlled clinical trials. Forty-four patients with definite RRMS, with two or more relapses in the previous two years, and a baseline EDSS < or = 5.5 were prospectively followed. The mean number of relapses spontaneously decreased from 1.72 (SD: 1.4) in the year prior to enrolment, to 1.0 (SD: 1.3) during the first year of follow-up (P < 0.05). Regression to the mean may explain as much as 40% of the reduction in the relapse rate from the baseline period to the period on-study.