The authors examined the impact of manipulative-based instruction on 2 independent cohorts of preservice elementary teachers. In Study 1, 50 participants engaged in problem solving with operations on whole numbers and fractions using concrete and representational manipulatives over 5 classes. Pre- to posttest performance on a mathematics survey and division-of-fractions task showed significant improvement in arithmetic knowledge, accompanied by a significant decrease in arithmetic misconceptions. Study 2 replicated Study 1 findings with an independent sample of 39 volunteers. These results indicate that manipulatives can effectively and efficiently reverse long-standing arithmetic misconceptions and increase arithmetic knowledge before education majors enter classrooms as full-time teachers. (Contains 4 tables and 1 figure.)