What is the nature of the association between age (7-11 years) and performance on mathematical equivalence problems (e.g., 7 + 4 + 5 + 7 + _)? Many prevailing theories suggest that there should be a positive association. However, change-resistance accounts (e.g., N. M. McNeil & M. W. Alibali, 2005b) predict a U-shaped association. The purpose of the present research was to test these differing predictions. Results from two studies supported a change-resistance account. In the first study (N = 87), performance on equivalence problems declined between the ages of 7 and 9 and improved between the ages of 9 and 11. The decrements in performance between the ages of 7 and 9 were then replicated in a second study (N = 35). Results suggest that the association between age and performance on equivalence problems is U-shaped.