Considerable research has addressed which “single” type of classroom goal structure influences students’ learning behaviors instead of multiple classroom goal structures. In this study, the impacts of approach-mastery goal structure and multiple goal structure on self-regulated learning strategies (SRL) were compared and changes of these impacts over time wereinvestigated with a quasi-experimental approach. Six classes with 152 [WM[1] seventh-grade junior high school students in Taipei City participated in this study.  All the classes were conducted as individual study groups for different manipulations,including 3 distinct levels: approach-mastery classroom goal structure, multiple classroom goal structure, and control group. The effects of classroom goal structures on SRL were examined at 3 stages: pre-test, post-test and delayed post-test. The results showed that there were significant two-way interactive effects between the classroom goal structures and measurement occasions on motivational/affective regulation (MR) and contextual regulation (CR) strategies. We found that multiple classroom goal structures had the best impact on students’ adaption of SRL because of its positive effects on MR and CR for both the post-test and delayed post-test; approach-mastery goal structure had minor positive effects on MR and CR because of its positive effects for only the post-test and not delayed post-test; the average scores for MR and CR in the control group declined or showed low-levels with time. In the conclusion, the implication for the theory and research are discussed.