Radial Frequency (RF) patterns are a useful stimulus for assessing sensitivity to changes in shape. With these patterns it is possible to separate sensitivity to local curvature information from the ability to globally integrate information around the contour. Previous work has demonstrated that young, school-age children are less sensitive to deformation in RF patterns than adults. However, since the efficiency of contour integration was not assessed, age-related differences in performance could arise from either changes in the strength of global pooling of information, the sensitivity to local curvature information, or both. In this study, psychophysical methods were used to reassess changes in sensitivity to RF patterns, separating changes in sensitivity to local curvature information from changes in contour integration strength. Typically developing observers (aged 6–24, N = 104) were tested using a 2-alternative forced-choice discrimination task with either 1, 2 or 3 cycles of sinusoidal modulation in a pattern of fixed RF. Thresholds were lower for older observers but the rate of change, as more modulation cycles were added, was approximately constant across age. The results indicate that changes in sensitivity to RF3 patterns across age are due to changes in local curvature sensitivity and not to the strength of contour integration.