Pre-cooling is used by many athletes for the purpose of reducing body temperature prior to exercise and, consequently, decreasing heat stress and improving performance. Although there are a considerable number of studies showing beneficial effects of pre-cooling, definite conclusions on the effectiveness of pre-cooling on performance cannot yet be drawn. Moreover, detailed analyses of the specific conditions under which pre-cooling may be most promising are, so far, missing. Therefore, we conducted a literature search and located 27 peer-reviewed randomized controlled trials, which addressed the effects of pre-cooling on performance. These studies were analysed with regard to performance effects and several test circumstances (environmental temperature, test protocol, cooling method, aerobic capacity of the subjects). Eighteen studies were performed in a hot (>26°C) environment and eight in a moderate. The cooling protocols were water application (n = 12), cooling packs (n = 3), cold drinks (n = 2), cooling vest (n = 6) and a cooled room (n = 4). The following different performance tests were used: short-term, high-intensity sprints (n = 2), intermittent sprints (n = 6), time trials (n = 10), open-end tests (n = 7) and graded exercise tests (n = 2). If possible, subjects were grouped into different aerobic capacity levels according to their maximal oxygen consumption (VO(2max)): medium 55-65 mL/kg/min (n = 11) and high >65 mL/kg/min (n = 6). For all studies the relative changes of performance due to pre-cooling compared with a control condition, as well as effect sizes (Hedges' g) were calculated. Mean values were weighted according to the number of subjects in each study. Pre-cooling had a larger effect on performance in hot (+6.6%, g = 0.62) than in moderate temperatures (+1.4%, g = 0.004). The largest performance enhancements were found for endurance tests like open-end tests (+8.6%, g = 0.52), graded exercise tests (+6.0%, g = 0.44) and time trials (+4.2%, g = 0.44). A similar effect was observed for intermittent sprints (+3.3%, g = 0.43), whereas performance changes were smaller during short-term, high-intensity sprints (-0.5%, g = 0.03). The most promising cooling methods were cold drinks (+15.0%, g = 1.68), cooling packs (+5.6%, g = 0.70) and a cooled room (+10.7%, g = 0.49), whereas a cooling vest (+4.8%, g = 0.31) and water application (+1.2%, g = 0.21) showed only small effects. With respect to aerobic capacity, the best results were found in the subjects with the highest VO(2max) (high +7.7%, g = 0.65; medium +3.8%, g = 0.27). There were four studies analysing endurance-trained athletes under time-trial conditions, which, in a practical sense, seem to be most relevant. Those studies found an average effect on performance of 3.7% (g = 0.48). In summary, pre-cooling can effectively enhance endurance performance, particularly in hot environments, whereas sprint exercise is barely affected. In particular, well trained athletes may benefit in a typical competition setting with practical and relevant effects. With respect to feasibility, cold drinks, cooling packs and cooling vests can be regarded as best-practice methods.