Physical activity and physical fitness are complex entities comprising numerous diverse components that present a challenge in terms of accurate, reliable measurement. Physical activity can be classified by its mechanical (static or dynamic) or metabolic (aerobic or anaerobic) characteristics and its intensity (absolute or relative to the person's capacity). Habitual physical activity can be assessed by using a variety of questionnaires, diaries, or logs and by monitoring body movement or physiologic responses. Selection of a measurement method depends on the purpose of the evaluation, the nature of the study population, and the resources available. The various components of physical fitness can be assessed accurately in the laboratory and, in many cases, in the field by using a composite of performance tests. Most coaches and high-level athletes would accept as very beneficial a dietary supplement that would increase performance in a competitive event by even 3%; for example, lowering a runner's time of 3 min, 43 s in the 1500 m by 6.7 s. To establish that such small changes are caused by the dietary supplement requires carefully conducted research that involves randomized, placebo-controlled, double-blind studies designed to maximize statistical power. Statistical power can be increased by enlarging sample size, selecting tests with high reliability, selecting a potent but safe supplement, and maximizing adherence. Failure to design studies with adequate statistical power will produce results that are unreliable and will increase the likelihood that a true effect will be missed.