A province wide prevalence study on multiple sclerosis (MS) was conducted in British Columbia (B.C.). The prevalence date was July 1, 1982. The major portion of this study was a review of all the files of neurologists practicing in B.C. as this was judged to be the most accurate source for identifying MS patients. 239,412 neurologists' files were hand searched by one researcher using modified Schumacher criteria for classification. Other sources used during the study for identifying MS patients were the MS Clinic, general practitioners, ophthalmologists, urologists, specialized facilities such as long term care facilities and rehabilitation centres, and patient self-referrals. A total of 4,620 non-duplicated cases were identified and classified. 4,112 of these (89%) were classified according to information contained in neurologists' records. The prevalence estimate for definite/probable MS in B.C. was 93.3/100,000 population. This increased to 130.5/100,000 population if possible MS and optic neuritis were also included. These rates are among the highest reported in Canada or elsewhere. The cooperation of B.C. neurologists made this study unique in its scope and accuracy of diagnosis.