The aim of this study was to investigate the effects of modelling, collaborative and game-based learning on geometry success in third-grade students. These approaches were applied to geometry instruction in nature on the success of students in geometry. The students' views about geometry activities were also examined. Explanatory design, one of the mixed methods in which qualitative and quantitative methods are both used, was used in the study. The study used a randomized pretest-posttest control group design for the quantitative data; phenomenological study, one of the qualitative research designs, provided the study's qualitative dimension. The study group consisted of 101 third-grade students attending three different public primary schools. There were 65 participants in the experimental group and 36 in the control group. The quantitative data were collected by a geometry success test and the qualitative data were collected through a structured interview form. The quantitative findings obtained at the end of the study revealed that the students' success in geometry was greatest in the modelling group. Qualitative findings showed that geometry activities in nature were more effective than in-class activities.