Multiple sclerosis (MS) is a devastating disease that can occur in early life, progressing to rapid disability and loss of physical, psychosocial and economic functioning, significantly affecting quality of life. The traditional treatment for MS has been symptomatic, treating acute relapses without affecting the underlying disease. The introduction of interferon-beta (IFN beta) has offered significant clinical benefits by reducing the frequency of relapses and slowing disease progression. Although the costs of this treatment are high, the costs to society of caring for a patient disabled by MS are greater, and if IFN beta can delay disease progression in the longer term, the economic impact would be substantial. Previous pharmacoeconomic studies of IFN beta have suggested that benefits can only be achieved at extremely high cost, with reported cost-effectiveness measures of up to 1 million pounds sterling (Pound) per quality-adjusted life year (QALY) [1995 values]. However, these studies have considered only the short term benefits of IFN beta treatment: over 2 to 3 years, the impact of treatment on patients' quality of life is relatively small, and cost-utility analyses that do not consider longer term benefits nor include societal costs may be misleading. The model reported here is based on the hypothesis that the delay in disease progression seen in short term clinical trials is likely to continue if treatment is continued. The model also assumes that the delay in disease progression, which represents a reduction in brain atrophy, will result in lasting clinical benefits even if treatment is stopped. These assumptions are strongly supported by clinical trial data and the treatment hypothesis itself. A delay in disease progression will result in a significant improvement in functioning and quality of life, and if the costs associated with increased disability can be postponed, even long term treatment of MS with IFN beta can be shown to be cost effective. Using resource utilisation costs derived from an economic evaluation of MS in the UK, it was possible to calculate the impact of delaying disease progression in terms of both health service and societal costs. An estimate of mean disease progression in patients with MS treated with IFN beta-1a compared with patients who did not receive disease-modifying agents suggested that significant cost savings would be realised after about 12 years' treatment with IFN beta-1a. The application of utility scores to the disease progression curves also facilitated estimates of cost effectiveness, with cost per QALY values ranging from 27,036 Pounds after 2 years' treatment with IFN beta-1a to 37,845 Pounds after 20 years' treatment (1995 values).