We construct a large panel dataset of schools and districts in Florida to evaluate curricular effectiveness in elementary mathematics. A key innovation of our study is that we allow for curriculum quality to be non-uniform across various mathematics subtopics. We find evidence of variability in curricular effectiveness across different subtopics "within" the same curriculum. Our findings suggest that educational administrators should consider the topical performance of their various curricular alternatives when making adoption decisions. (Contains 2 figures and 9 tables.)