To resolve some of the controversy regarding insulin regulation of blood flow, we performed in 20 normal subjects a) a reproducibility study of plethysmographic, Doppler ultrasound and laser Doppler blood flow measurements (n = 7), b) a sequential insulin dose-response study with measurement of forearm (plethysmography), leg (Doppler ultrasound) and skin (laser Doppler) blood flow (n = 12), and c) a sequential insulin dose-response study with comparison of forearm (plethysmography) and calf (plethysmography) blood flow (n = 8). We also searched for factors which might explain the interindividual variation in the blood flow response to insulin. During sequential insulin infusions (2 h each, 61 +/- 2, 139 +/- 6, 462 +/- 15 mU/l), forearm blood flow increased by 17 +/- 6, 50 +/- 14 and 113 +/- 17% (p < 0.05 or less between steps), respectively. The increase at the 61 +/- 2 mU/l insulin concentration barely exceeded methodological variation (13 +/- 2%). In contrast to the continuous increase in blood flow, the glucose arterio venous difference reached its maximum (1.7 +/- 0.2 mmol/l) at the lowest 61 +/- 2 mU/l insulin concentration and remained constant thereafter. Forearm and calf blood flow responses to insulin were virtually identical when determined with plethysmography. In contrast, only a 27% increase was detected in femoral flow index as determined by Doppler ultrasound. Forearm blood flow (per forearm volume) was highly correlated with the relative forearm muscle content (mean 59 +/- 5%, range 24-81%) both basally (r = 0.86, p < 0.001, n = 12) and at all insulin concentrations (r = 0.85-0.92, p < 0.001) indicating that the percent of forearm that is muscle explains 70-85% of interindividual variation in blood flow. In conclusion 1) physiological insulin concentrations stimulate glucose uptake mainly by increasing glucose extraction while supraphysiological insulin concentrations increase forearm glucose uptake predominantly via increases in blood flow. 2) The dose-response characteristics of insulin stimulation of forearm and calf blood flow are similar when determined with strain-gauge plethysmography. 3) Relative forearm muscle content is a key factor in determining both basal forearm blood flow and the interindividual variation in its response to insulin in normal subjects.