We set out to evaluate a clinical foot-screening programme in terms of primary outcomes (reductions in the incidence of ulcers and lower limb amputation) and process outcomes (compliance with screening, the number of patients not completing the programme and the use of chiropody services and prescribed footwear and cost). All but 4 of 2001 patients attending a general diabetic out-patient clinic were allocated randomly to index and control groups. The exceptions were patients who presented with active ulcers and were placed in the index group. Primary and secondary screening programmes identified 128 high risk patients in the index group and these were admitted to the foot protection programme. At 2-year follow-up, 11 fewer ulcers were reported from the index group. There were 7 amputations (1 major, 6 minor) in the index group and 23 (12 major and 13 minor) in the control group. The differences were not statistically significant for ulceration or minor amputations but significant for major amputations (p < 0.01). The total cost of the 2-year programme was pounds sterling 100,372 (1991-92 costs), with a mean cost per patient of approximately pounds sterling 100. Taking pounds sterling 12,000 as a conservative estimate of the cost of a major amputation, the foot clinic was cost-effective in terms of amputations averted. The process outcomes were much less satisfactory. Cost-effectiveness could have been improved if it had been possible to improve patient compliance.