While warm up is considered to be essential for optimum performance, there is little scientific evidence supporting its effectiveness in many situations. As a result, warm-up procedures are usually based on the trial and error experience of the athlete or coach, rather than on scientific study. Summarising the findings of the many warm-up studies conducted over the years is difficult. Many of the earlier studies were poorly controlled, contained few study participants and often omitted statistical analyses. Furthermore, over the years, warm up protocols consisting of different types (e.g. active, passive, specific) and structures (e.g. varied intensity, duration and recovery) have been used. Finally, while many studies have investigated the physiological responses to warm up, relatively few studies have reported changes in performance following warm up. The first part of this review critically analyses reported changes in performance following various active warm-up protocols. While there is a scarcity of well-controlled studies with large subject numbers and appropriate statistical analyses, a number of conclusions can be drawn regarding the effects of active warm up on performance. Active warm up tends to result in slightly larger improvements in short-term performance (<10 seconds) than those achieved by passive heating alone. However, short-term performance may be impaired if the warm-up protocol is too intense or does not allow sufficient recovery, and results in a decreased availability of high-energy phosphates before commencing the task. Active warm up appears to improve both long-term (>/=5 minutes) and intermediate performance (>10 seconds, but <5 minutes) if it allows the athlete to begin the subsequent task in a relatively non-fatigued state, but with an elevated baseline oxygen consumption (VO(2)). While active warm up has been reported to improve endurance performance, it may have a detrimental effect on endurance performance if it causes a significant increase in thermoregulatory strain. The addition of a brief, task-specific burst of activity has been reported to provide further ergogenic benefits for some tasks. By manipulating intensity, duration and recovery, many different warm-up protocols may be able to achieve similar physiological and performance changes. Finally, passive warm-up techniques may be important to supplement or maintain temperature increases produced by an active warm up, especially if there is an unavoidable delay between the warm up and the task and/or the weather is cold. Further research is required to investigate the role of warm up in different environmental conditions, especially for endurance events where a critical core temperature may limit performance.