Regular exercise is important for patients with multiple sclerosis (MS) to maintain their functional ability and general health. The aim of this study was to determine whether a long-term exercise program has any effect on functional impairment or healthrelated quality of life (HRQOL) in subjects with mild to moderate MS. In a randomised controlled trial, subjects in the intervention group (n = 47) exercised according to a progressive exercise program, mainly consisting of resistance training, for six months. Subjects in the control group (n = 48) received no intervention. The subjects were assessed at baseline and at six months using the Multiple Sclerosis Functional Composite (MSFC), the Expanded Disability Status Scale (EDSS), the Functional Independence Measure (FIM), the MS Quality of Life-54 (MSQOL-54) questionnaire and the Centre for Epidemiologic Studies Depression Scale (CES-D). The drop-out rate was low (4%) with 91 subjects completing the study. At six months, the exercising subjects showed improvement on the MSFC (mean score change 0.114, 95% confidence interval [CI] 0.010 to 0.218), whereas the control subjects showed deterioration (mean score change -0.128, 95 % CI -0.232 to -0.025). The change between groups was statistically significant (interaction, p = 0.001). Consistent with the physical nature of the intervention, the change predominantly occurred in leg function/ambulation. The effect seen in the EDSS, FIM, MSQOL-54 or CES-D was nil. These findings indicate that MSFC is more sensitive than EDSS in the detection of improvement in functional impairment as a result of regular exercise. The unfavourable results from HRQOL do not rule out the possibility that other types of exercise programs may improve it in MS.