The hypothesis that hyperperfusion decreases muscle fatigue by increasing O2 and substrate delivery to the muscle was tested. Canine gastrocnemius-plantaris in situ preparations were stimulated at 5 Hz for 4 min during a free-flow control period and for 20 min during a pump-perfused experimental period. O2 delivery during these two periods was matched either by decreasing blood flow in animals breathing 100% O2 (high O2/low flow) [experimental-to-control ratio (E/C) = 0.97 + 0.02] or by increasing the blood flow in animals breathing 14% O2 (low O2/high flow) (E/C = 1.01 + 0.01). Plasma flow estimated from hematocrit to approximate substrate delivery was matched in the two contraction periods either by maintaining blood flow at the steady-state level (constant flow) (E/C = 0.98 + 0.10) or by increasing flow in animals with a dextran for 6% of blood volume exchange (dilute/high flow) (E/C = 1.02 + 0.02). E/C for initial developed tension was 1.00 + 0.02. Over 20 min, developed tension decreased 15.0 + 1.1% with low O2/high flow and 16.0 + 1.8% with dilute/high flow. Tension decreased by 28.0 + 3.0 and 27.8 + 1.5% with high O2/low flow and constant flow, respectively. Thus hyperperfusion decreased fatigue by a mechanism independent of increased O2 and substrate delivery.