A simple explanation for what determines the rate of energy consumption of running animals was proposed and tested. Two questions were specifically addressed: why does the rate of energy consumption increase linearly with running speed and why is running more expensive for smaller animals. It was hypothesized that the rate of energy consumption is set by the cost of generating muscular force to support body weight. In a running animal, most of the force is exerted vertically on the ground and this force averaged over a complete stride is equal to body weight. At faster running speeds and in smaller animals the feet are in contact with the ground for shorter periods of time. This presumably involves the recruitment of faster muscle fibers which consume more energy per unit force. Thus, it was hypothesized that the rate of energy consumption,(Èmetab) per unit body weight (Wb) would be inversely proportional to the time each foot is in contact with the ground (tc), that is, Emetab/Wb = c 1/tc proportionality factor, c can be considered a "cost coefficient". This hypothesis was tested on five species of bipedal hopping and quadrupedal mammals ranging in size from 40-gram kangaroo rats to 140-kg ponies. Across the entire aerobic speed range, metabolic rate increased by more than 300% but the cost coefficient varied by less than 25%. Across the 3,500-fold size range the cost coefficient was also nearly invariant, scaling with body weight to only the 0.04 power. Data from bipedal running birds and humans was also analyzed in this way with similar results except that the cost coefficients were substantially greater. Thus, this simple explanation for the rate of energy consumption by running animals appears to work remarkably well for a wide variety of animals.