This study investigated the transfer of solution strategies between close variants of quantitative reasoning questions. Pre- and posttests were obtained from 406 college undergraduates, all of whom took the same posttest; pretests varied such that one group of participants saw close variants of one set of posttest items while other groups saw close variants of other sets. Some participants also saw items that looked similar to some posttest questions but were mathematically different. Between pre- and posttests, some participants viewed solution rationales explaining their incorrect answers. Students at all ability levels performed better on close variants, suggesting there was some positive transfer, but the presence of appearance variants interfered with this transfer. This study suggests an important first step toward understanding how tests and item vats can be designed to capitalize on the extent to which students set up test-taking strategies.