The aim of this investigation was to develop a test battery to measure the motor abilities of 4- and 5-year-old children. The subjects were 932 healthy kindergarten children from Tartu, Estonia: boys, 4 years, n = 205; boys, 5 years, n = 267; girls, 4 years, n = 224; girls, 5 years, n = 236. Height and body mass were measured, and the body mass index (BMI, kg/m ) was calculated. The following motor ability tests were used: 3-min run (cardiorespiratory endurance), standing long jump (explosive strength), 4 × 10 m shuttle-run (running speed, agility), situps for 30 sec (trunk strength), sit-and-reach (flexibility), and sand bag (150 g) throw with the dominant hand (explosive strength, coordination). The motor tests were repeated 1 week later in subgroups of 28-33 children. Most of the tests had an acceptable reliability (r > 0.75) and none had a poor reliability (r < 0.40). There were significant (P < 0.05-0.001) differences between 4- and 5-year-old boys and girls in all tests except the sit-and-reach between the 4- and 5-year-old girls. Performances of boys were generally better than those of girls of the same age. Test results were often dependent on height and body mass, but not on the BMI. When body size and age were statistically controlled with second-order partial correlations, the relationships were greatly reduced or no longer significant. Stepwise regression analysis revealed that age, height, and mass of the children accounted for only a small percentage of the variance in performances. The greatest contribution to total variance was accounted for by body size in the 3-min run in 4-year-old boys (6%). Am. J. Hum. Biol. 9:659-664, 1997. © 1997 Wiley-Liss, Inc.