To study the effect of long-term treatment of the platelet inhibitor ticlopidine as secondary prevention against the need of vascular surgery in patients with intermittent claudication. The Swedish Ticlopidine Multicentre Study (STIMS), was conducted in six medical and surgical clinics of university hospitals in Sweden. 687 claudicants were randomised to ticlopidine 250 mg bd or placebo and vascular surgery events were recorded prospectively over a 7-year period. Cox proportional hazards models of risk for leg vascular surgery were constructed using drug treatment and 11 putative risk factors for vascular disease as covariates. Surgical event-free survivals were compared by Kaplan-Meier analysis. The overall rate of first operations was 2.4% per annum. More than half of these operations were in the aortoiliac region. One-quarter of patients operated during the period required further operations but amputation was rare. Ticlopidine treatment reduced the need for vascular reconstructive surgery by about half, both in intention-to-treat and on-treatment analyses (unadjusted relative risks 0.486, 95% CI 0.317-0.745; p < 0.001; 0.493, 95% CI 0.290-0.841: p < 0.01, respectively). In Cox model analysis only male sex was confirmed as a risk factor for surgery. Previous peripheral arterial surgery was the strongest predictor of the need for surgery. None of the risk factors examined interacted statistically with the effect of treatment with ticlopidine. In patients with intermittent claudication it seems possible to prevent the need for future vascular surgery by the use of platelet inhibition with ticlopidine.