Multiple sclerosis (MS) patients treated with interferon-beta (IFN-beta) often form anti-IFN-beta antibodies accompanied by a reduction in IFN-beta bioavailability. The clinical effect of these antibodies remains controversial. MS patients in British Columbia, Canada, must be diagnosed and evaluated annually by neurologists in an MS clinic in order to be reimbursed for their IFN-beta prescriptions. We have identified at the UBC MS clinic a cohort of 262 patients, each having been treated with a single IFN-beta preparation more than three years, some for nearly a decade. Of 119 patients treated with Betaseron (R) (IFN-beta 1 b), 18 (15.1 %) were neutralizing antibody positive (NAb+) at the time of the study, whereas of 131 treated with subcutaneous Rebif (R) (IFN-beta 1 a SC), 16 (12.2%) were NAb+, but none of 12 treated with intramuscular Avonex (R) (IFN-beta 1 a) had detectable neutralizing antibodies. During the first two years of treatment, the relapse rate was significantly reduced from pre-treatment rates (P < 0.001) and appeared to be unaffected by the subsequent NAb status. However, the relapse rates in the NAb+ patients were significantly greater than in the NAb- patients during years 3 (P < 0.010) and 4 (P < 0.027). Betaseron (R)-treated NAb+ patients tended to have more relapses than NAb- patients during year 3 and this almost reached significance (P = 0.056) but their relapse rate did not differ in year 4 and later. In contrast, Rebif (R)-treated NAb+ patients tended to have more relapses in year 3 than Rebif (R)-treated NAb- patients (P = 0.074), but in year 4 they clearly (P = 0.009) had more relapses than Rebif (R)-treated NAb- patients. There was no convincing effect on progression of disability in any group.