OBJECTIVES: To determine time trends and geographical variations in the incidence of major amputation for peripheral arterial disease and whether lower rates of amputation were related to higher rates of arterial reconstruction. DESIGN: Analysis of Scottish hospital discharge data. SETTING: Scotland 1981-1990. MATERIALS: Patients undergoing major amputation or arterial reconstruction for peripheral arterial disease. CHIEF OUTCOME MEASURES: Time trends in age-sex standardised rates of major amputation and arterial reconstruction, and correlation between the rates of these operations by health board. MAIN RESULTS: In Scotland, between 1981 and 1990, the incidence of major amputation fell by 22% (p < 0.001). Inconsistencies were observed within different age-sex groups. In the population under 65 years of age the incidence of amputation fell by 45% (p < 0.001), whereas in those over 65 years the incidence increased by 54% (p < 0.001). Amputation rates fell in men but a paradoxical increase was observed in women. Between 1981 and 1990, rates of arterial reconstruction doubled (p < 0.001), with an increase in all age-sex groups. Rates of amputation and reconstruction varied between health boards of residence, with a positive correlation (r = 0.5) between rates of operations within health board. Therefore areas with higher reconstruction rates tended to have higher amputation rates. CONCLUSIONS: In Scotland, the incidence of amputation has fallen during a period when reconstruction rates have risen greatly. However inconsistencies in time trends by age-sex groups, and the lack of an inverse correlation by health board of residence, suggest that fewer amputations are unlikely to be due solely to an increase in reconstructive surgery.