This study has investigated the effects of problem solving instruction on physics achievement, problem-solving performance and strategy use in an introductory physics course at university level. In this study, pretest-posttest and quasi-experimental design with nonequivalent control group was used. Two groups of student teachers (n=74) participated in this study. During the 8-week study, one group received the strategy instruction while the other group acted as control. Data of the study were collected by Physics Achievement Test (PAT), Problem-Solving Performance Test (PSPT) and Problem-Solving Strategies Scale (PSSS). Findings of the study indicate that strategy instruction was effective on physics achievement, problem-solving performance, and strategy use. The implications of these results for physics instruction are discussed.