The purpose of this study was to validate a two-dimensional (2D) echography coupled range-gated Doppler system for the non-invasive measurement of internal diameter, blood flow velocity, and pulse wave velocity of peripheral arteries, such as the common carotid artery (CCA), femoral artery (FA), and brachial artery (BA) in man. The array of the ultrasonic system and the Doppler probe were attached and formed a fixed angle (38 degrees 30'). The artery was firstly visualized using the echo-graphic array probe in order to position the Doppler beam. Then, the range-gated Doppler system was used to measure both internal diameter and blood flow velocity with the sample volume position covering the internal diameter. Using a hydraulic device, there was an obvious correlation between the calculated and the measured velocities (r = 0.98). Normal values of diameter, blood flow velocity and blood flow were measured in 18 healthy volunteers. The means (+/- 1 standard deviation) was as follows: diameter, CCA = 0.636 +/- 0.027 cm, FA = 0.843 +/- 0.074 cm, BA = 0.302 +/- 0.052 cm; flow velocity, CCA = 19.5 +/- 2.1 cm s-1, FA = 11.4 +/- 1.2 cm s-1, BA = 6.7 +/- 1.0 cm s-1. Blood flows were as follows: CCA, 370.6 +/- 42.5 ml mn-1, FA 387.0 +/- 75.0 ml mn-1 and BA (wrist occlusion) 29.8 +/- 12.5 ml mn-1. The intra-observer reproducibilities for CCA, FA and BA were respectively: for diameter, 4.9%, 4.12% and 14.8%; for velocity, 8.9%, 10.6% and 10.2%. The inter-observer reproducibilities were respectively: for diameter, 5.6%, 5.4% and 11.3% for velocity, 6.5%, 5.7% and 6.3%. Simultaneous determinations of pulse wave velocity from blood flow velocity recording allowed estimations of the distensibility of these arteries. Finally, the coupled echo and range-gated Doppler system permitted non-invasive evaluation of blood flow calculated as the product of the vessel cross-sectional area and measured blood velocity and also of arterial compliance as the ratio of the cross-sectional area and the square of pulse wave velocity. Such estimations of the conduit and buffering functions of peripheral large arteries in man were shown to be more accurate for the common carotid and the femoral arteries than for the brachial artery.