Performance differences in dynamic and static balance ability of 150 preschool Ss aged 3,4, and 5 yr. were studied. Ss performed 4 balance-beam tasks and 2 balance-board tasks. An Age by Sex (3 × 2) factorial, design employing both univariate and multivariate ANOVA techniques were the statistics used. For both dynamic and static balance Age was highly significant, and the use of multivariate ANOVA indicated significant sex differences on the static balance tasks. The appropriateness of multivariate techniques where more than one dependent variable is measured on the same population was discussed, and the need to take into account the relationship between these variables when analyzing the data was noted.