The present study compared the microdialysis ethanol outflow-inflow technique for estimating blood flow (BF) in skeletal muscle of humans with measurements by Doppler ultrasound of femoral artery inflow to the limb (BFFA). The microdialysis probes were inserted in the vastus lateralis muscle and perfused with a Ringer acetate solution containing ethanol, [2-3H]adenosine (Ado), and D-[14C(U)]glucose. BFFA at rest increased from 0.16 +/- 0.02 to 1.80 +/- 0.26 and 4.86 +/- 0.53 l/min with femoral artery infusion of Ado (AdoFA,i) at 125 and 1,000 microg . min-1 . l-1 thigh volume (low dose and high dose, respectively; P < 0.05) and to 3.79 +/- 0.37 and 6.13 +/- 0.65 l/min during one-legged, dynamic, thigh muscle exercise without and with high AdoFA,i, respectively (P < 0.05). The ethanol outflow-to-inflow ratio (38.3 +/- 2.3%) and the probe recoveries (PR) for [2-3H]Ado (35.4 +/- 1.6%) and for D-[14C(U)]glucose (15.9 +/- 1.1%) did not change with AdoFA,i at rest (P = not significant). During exercise without and with AdoFA,i, the ethanol outflow-to-inflow ratio decreased (P < 0.05) to a similar level of 17.5 +/- 3.4 and 20.6 +/- 3.2%, respectively (P = not significant), respectively, while the PR increased (P < 0.05) to a similar level (P = not significant) of 55.8 +/- 2.8 and 61.2 +/- 2. 5% for [2-3H]Ado and to 42.8 +/- 3.9 and 45.2 +/- 5.1% for D-[14C(U)]glucose. Whereas the ethanol outflow-to-inflow ratio and PR correlated inversely and positively, respectively, to the changes in BF during muscular contractions, neither of the ratio nor PR correlated to the AdoFA,i-induced BF increase. Thus the ethanol outflow-to-inflow ratio does not represent skeletal muscle BF but rather contraction-induced changes in molecular transport in the interstitium or over the microdialysis membrane.