Available estimates of the frequency with which a patient with optic neuritis develops multiple sclerosis range from as low as 13% to as high as 87%. In an effort to obtain a better estimate, a nation-wide study of optic neuritis was carried out in Israel. Patients who fulfilled strict diagnostic criteria of optic neuritis were identified and examined periodically. Between 1955 and 1964, 105 patients were found and on the basis of these, the average annual age-adjusted incidence of optic neuritis in Israel was 0.56 per 10(5) population compared to 1.2 per 10(5) cases of multiple sclerosis per year, i.e. optic neuritis was about half as frequent as multiple sclerosis each year. As with multiple sclerosis, optic neuritis was more common in European immigrants to Israel than Afro-Asian immigrants. During a follow-up interval which ranged from 3.3 to 15.6 years (mean 9.5 years), at least 27 of the 105 patients developed multiple sclerosis (28%). A life-table analysis showed that after 10 years 32.3 +/- 5.6% of patients with optic neuritis would develop multiple sclerosis and, after 14 years, about half would develop multiple sclerosis. Risk of dissemination was highest in those who were youngest when optic neuritis developed. Neither sex nor ethnic background influenced risk significantly. Results of the present study support earlier work using life-table methods carried out in Hawaii which also showed that between 29 and 39% of patients with optic neuritis will develop multiple sclerosis within 10 years of onset. The life-table method is a better predictor of prognosis than newer laboratory techniques such as spinal fluid studies of IgG, kappa-lambda light chain ratios and serum/CSF IgG ratios.