OBJECTIVES: to estimate the intra-observer variability of the measurement of the ankle-brachial systolic pressure index (ABPI) and to compare the reproducibility of the measurements by experienced vascular laboratory assistants and by less-experienced general practice personnel. DESIGN: repeated measurement of ABPI by general practitioners (GPs), GP-assistants and vascular laboratory assistants using a pocket Doppler device and a random-zero sphygmomanometer. METHODS AND MATERIALS: ABPI was measured in six patients with various degrees of PAOD by two experienced observers (vascular laboratory assistants) and by 24 less-experienced observers (18 practice assistants, six GPs). RESULTS: the total number of measurements was 354. The overall intra-observer variability estimate was 11.8% ABPI. The intra-observer variability was 7.3% in the experienced observers and 12.0% in the less-experienced observers. The difference of variability between experienced and less-experienced observers was significant. CONCLUSIONS: the ABPI is suitable in follow-up studies where repeated measurements are needed. Differences between measurements can be minimised by performing repeated measurements or by using more experienced observers. Copyright 1999 W.B. Saunders Company Ltd.