Mathematical inquiry involving hands-on activities has received increasing attention in mathematics education. Besides various customized physical teaching aids, subject-specific information technology, such as Dynamic Geometry System (DGS), finds extensive use in mathematical inquiry activities. However, effects of DGS and physical manipulatives on inquiry-based math learning remain an open question. Hence, by adopting a quasi-experimental research design, this paper aims to empirically compare the immediate learning outcomes, knowledge retention, and learning interest of seventh-grade students who explore with virtual manipulatives (i.e., the DGS) and who explore with physical manipulatives. Specifically, 131 students participated in learning activities centered on exploring pyramids and prisms. During the inquiry process, Group A (n = 33) constructed pyramids and prisms in DGS, Group B (n = 34) observed pre-made virtual models in DGS, Group C (n = 32) observed physical models, and Group D (n = 32) made physical pyramids and prisms with polymer clay and small sticks. Moreover, pretest, post-test, and delayed post-test designed according to the Van Hiele model, as well as an adapted interest questionnaire, were employed to evaluate students' performances; and collected data were analyzed by means of ANCOVA and t-test. Findings of the study revealed that in the context of construction, students employing the DGS exhibited superior immediate learning outcomes and greater knowledge retention compared to their peers who utilized physical manipulatives; while in the case of observation, the virtual and physical manipulatives yielded similar impacts on students' immediate learning outcomes, but students who involved in the DGS demonstrated higher knowledge retention. Furthermore, regarding the DGS environment, students who engaged in constructive manipulation surpassed their peers who engaged in observing manipulation. In terms of interest, the DGS proved to be more effective in both stimulating and maintaining higher interest compared to the physical manipulatives, and constructive manipulation was more effective than observing manipulation. In summary, as compared to the physical manipulatives, DGS, particularly when employed with constructive strategy, has shown to encourage students to more actively engage in task-related cognitive behaviors, thereby supporting and enhancing inquiry-based math learning of students in junior high school.