It is well established that for work requiring high power output, endurance time is short, and that low power outputs can be maintained for long periods. Parameters describing this relationship are important in characterising work performance and the capacity of humans as a source of mechanical power. The purpose of this paper is to provide a brief review of the available literature investigating this relationship and its parameters. Most experimental data reflect measurements of endurance times over a range of constant power outputs on the cycle ergometer. Early graphical analyses of these data have been superseded by curve fitting, which in turn has led to establishment of the two component hyperbolic model now embodied in the critical power test. This model has been modified and extended in various ways to account for its shortcomings. In addition, a number of different exercise forms have been studied, and the effects of a variety of secondary factors (training status, age, sex, for example) on the parameters have also been investigated.