Compression garments (CGs) provide a means of applying mechanical pressure at the body surface, thereby compressing and perhaps stabilizing/supporting underlying tissue. The body segments compressed and applied pressures ostensibly reflect the purpose of the garment, which is to mitigate exercise-induced discomfort or aid aspects of current or subsequent exercise performance. Potential benefits may be mediated via physical, physiological or psychological effects, although underlying mechanisms are typically not well elucidated. Despite widespread acceptance of CGs by competitive and recreational athletes, convincing scientific evidence supporting ergogenic effects remains somewhat elusive. The literature is fragmented due to great heterogeneity among studies, with variability including the type, duration and intensity of exercise, the measures used as indicators of exercise or recovery performance/physiological function, training status of participants, when the garments were worn and for what duration, the type of garment/body area covered and the applied pressures. Little is known about the adequacy of current sizing systems, pressure variability within and among individuals, maintenance of applied pressures during one wear session or over the life of the garment and, perhaps most importantly, whether any of these actually influence potential compression-associated benefits. During exercise, relatively few ergogenic effects have been demonstrated when wearing CGs. While CGs appear to aid aspects of jump performance in some situations, only limited data are available to indicate positive effects on performance for other forms of exercise. There is some indication for physical and physiological effects, including attenuation of muscle oscillation, improved joint awareness, perfusion augmentation and altered oxygen usage at sub-maximal intensities, but such findings are relatively isolated. Sub-maximal (at matched work loads) and maximal heart rate appears unaffected by CGs. Positive influences on perceptual responses during exercise are limited. During recovery, CGs have had mixed effects on recovery kinetics or subsequent performance. Various power and torque measurements have, on occasions, benefitted from the use of CGs in recovery, but subsequent sprint and agility performance appears no better. Results are inconsistent for post-exercise swelling of limb segments and for clearance of myocellular proteins and metabolites, while effects on plasma concentrations are difficult to interpret. However, there is some evidence for local blood flow augmentation with compression. Ratings of post-exercise muscle soreness are commonly more favourable when CGs are worn, although this is not always so. In general, the effects of CGs on indicators of recovery performance remain inconclusive. More work is needed to form a consensus or mechanistically-insightful interpretation of any demonstrated effects of CGs during exercise, recovery or - perhaps most importantly - fitness development. Limited practical recommendations for athletes can be drawn from the literature at present, although this review may help focus future research towards a position where such recommendations can be made.