The course of multiple sclerosis was monitored by means of a long-term prospective study. This paper reports on first results from this study, covering 102 patients who were monitored for 2 years or more and examined regularly. It is attempted to quantify the extent of the worsening or improvement in the patients' clinical status over the period of observation using regression analysis techniques. The severity of the disease did not correlate to the period of observation in 15% of the cases, these displaying no progression with regard to the clinical signs. In most cases (32%) a description of the disease progression was possible by means of a linear regression line, 21% had a parabolic and 23.5% an increasingly progressive course. For nine patients a second or third degree polynomial regression curve could be used to describe the course of the disease. The individual progression of the disease as estimated using the linear regression coefficient did not correlate with the individual relapse rate (on average 1.1 per year), nor with the age of the patient, the severity of the multiple sclerosis, the duration of the illness or with the previous course of the disease.