This study investigated how to present grammatical structures to intermediate-level French college students. It compared the effects of a guided inductive and a deductive approach on short- and long-term learning of 10 structures. A mixed-methods design was adopted to assess learning of the structures and to investigate preference of approach. Performances in both conditions were measured through a within-subjects design featuring a pretest/posttest and immediate posttreatment tests. A questionnaire assessed students' preferences and relationships between preferences and performance were examined. Findings indicated a significantly greater effect of the guided inductive approach on short-term learning. The long-term findings and the relationship between preferences and performances were not significant. Analyses indicated that students who preferred explanations of the rules performed better with a guided inductive approach. (Contains 8 tables.)