The maintenance of adequate oxygen delivery (DO2) and tissue uptake (VO2) has become central dogma in the management of the critically ill. However, these parameters are derived using gas tensions measured in mixed venous blood and may not reflect changes in regional blood flow. Therefore, it has become necessary to provide estimates of blood flow to specific organs and to evaluate the most adequate techniques available. In order to define the best means of assessing blood flow to the lower limb noninvasively in normal subjects, measurements of superficial femoral arterial blood flow using Doppler ultrasound (DU) and strain gauge plethysmography (SGP) were compared in 10 normal volunteers at rest and during exercise. To evaluate the effect of strain gauge positioning, results of measurements made under four different combinations of cuff/strain gauge placement were compared in 15 other volunteers. The correlation of the limb blood how obtained using the two methods at rest and exercise was 0.57 and 0.62 and the limits of agreement (($) over bar d +/- 2SD) were 0.40 +/- 2.49 and - 0.86 +/- 5.22 ml 100 ml(-1) tissue min(-1) at rest and on exercise, respectively. Results obtained using SCP were more reproducible (Coef. repeat. 0.45 vs. 0.94 ml 100 ml(-1) tissue min(-1), for SGP and DU, respectively). The various combinations of cuff/strain gauge positioning showed a tendency to over-read when the latter was placed on the thigh, but were not significantly different (P > 0.05). Measurements of limb blood flow obtained using DU and SGP correlate poorly over a wide range of blood flow and do not agree, the results from the latter being more reproducible. Although the same position of cuff and strain gauge should be maintained throughout an experiment, varying the positions studied yields comparable results.