We tend to assume that natural selection will bring about 'optimal' configurations in morphology and behaviour. Jumping locomotion involves large forces and energy costs which, in this non-cyclic activity, are generated anew with each jump. Jumping appears to be, therefore, a major target for optimization. It has been a standard assumption that jumpers will tend to adopt ballistic paths which will minimize the energy costs involved in jumping, and will act to minimize the loads applied to the body. Experimental studies, using kinematic analysis of digitized video recordings of the jump in five prosimian primates, with a 25-fold range in body mass, show that most do not adopt energy-efficient paths until the length of the jump is close to the maximum they can attain. Statistical analysis of quantified field observations suggests that, of three primate jumpers, only the largest, most unspecialized appears to take the forces applied to the musculoskeletal system into consideration when selecting supports used in locomotion. 'Ecological' factors, such as time pressure and habitat support density, may thus be the prime consideration for many species in determining the manner in which they jump.