Purpose: Based on the affordances concepts, grasping ball with various sizes was designed to examine age differences in finger numbers and body-scaled ratios of prehension. Methods: Eighteen participants were in each of child (6-yr-old), adult (30-yr-old), and older adult (70-yr-old) groups (N = 54). All participants were asked to grasp a stationary ball (diameter are 2, 4, 6, 8, and 10 cm.) in separate trial and each size for 3 trials. Two researchers of motor behavior were employed in video systematic observation method for analyzing filmed prehensile trials. The inter-rater reliability was .80 and the intra-rater reliability were .88 and .91, respectively. Results: (1) χ^2 test showed that child group utilized the finger numbers more in relative bigger sizes ball than adult and older adult groups significantly (p ＜ .05); and (2) Mixed-design two-way ANOVA, 3 (age) × 5 (ball sizes) with the ball sizes were repeated measure, and Tukey's HSD Post Hoc test showed significant main effects (p ＜ .05) in age and ball sizes, that revealed the body-scaled conduced ratios of child group were significant (p ＜ .05) different with adult (R(subscript 0) = 1.80) and older adult (R(subscript 0) = 1.69) groups, and the bigger sizes ball had bigger body-scaled conduced ratios (R(subscript 1) = 0.79, R(subscript 2) = 1.35, R(subscript 3) = 1.73, R(subscript 4) = 2.03, R(subscript 5) = 2.32). Conclusions: (1) Children's prehension utilized more fingers for relative bigger ball than adults and older adults did; and (2) Children's body-scaled ratios of prehension smaller then adults and older adults, and there were age differences in body-scaled ratios of prehension.