Reports dating back to the 1920s have unequivocally established that efficiency of movement improves during infancy and childhood, and generally through adolescence. The inclination from 1940 to 1960 to record movement in quantitative values has given way during the last two decades to a reemphasis of qualitative assessment. Investigators commonly agree that movement skills change in an orderly manner, but controversy continues over the degree of accuracy obtainable with the various descriptive forms, and over the utility of products resulting from such assessments. Comparison of the quantitative changes in movement skills of children is difficult because standardized procedures of test administration have not been applied. When comparisons are possible because of similar or identical testing protocols, improvement in selected motor tests is evident in both boys and girls until adolescence. At approximately 13 years of age the performance of girls in some tests reaches a plateau, and may even decline thereafter, while boys continue to improve in skills requiring strength, power, and muscular endurance. Exceptions to these generalizations occurred in arm and shoulder girdle muscular endurance, as measured by the flexed-arm hang, where boys had superior performances beginning at age 7, and in flexibility, as measured by the sit and reach test, where girls excelled at age 5 and thereafter. Stability of motor performance was greater for tasks that required all-out effort than for those emphasizing accuracy or total body coordination. Relationships between successive measures taken during early and middle childhood are likely to decline more rapidly than those taken after adolescence. Girls generally were more stable in motor performance than boys, except in the Motor Performance Study, where the values for boys across a range of 5 to 6 years were clearly more stable. This review underscores the need for careful documentation of the conditions under which data on motor performance are obtained. Numerous reports were examined and excluded by the authors because information that would have qualified the data for comparative analysis was not available. Essential ingredients in such reports are descriptions of the sampling techniques and the manner of calculating chronological ages, socioeconomic status, ethnic and racial characteristics, evidence of secular changes, geographic and environmental characteristics, and a detailed account of the testing procedures. Differences in maturational age for a given chronological age, and accompanying assumptions about body size, may account for differences in motor skills that might otherwise be attributed to changes in p