Four experiments compared the re-solution performance of prior solvers with that of prior nonsolvers given the correct solutions. Experiments 1 and 2 challenged Weisberg and Alba's (1981) contention that solving a problem and being shown the solution yield equivalent problem knowledge. In both experiments, students who initially solved problems showed near-perfect recall of the solutions after a 1-week delay, far superior to recall by students who had been shown the correct answers. In Experiment 3, solves showed poor solution retention when the connection between the problem and the solution was not meaningful. Experiment 4 showed that with meaningful problems, solvers and those merely provided with solutions have qualitatively different problem representations. The findings can be explained in terms of differential understanding of problems and their solutions.