The purpose of this study is to identify the effect of using metacognitive strategies for problem solving in "geometry" on fifth grade students' achievement, metacognitive skills and attitude. Experimental method was used with a pretest/posttest control group design. Firstly, both groups were subject to a pretest that was comprised of "the achievement test", "the metacognitive skills inventory" and "the attitude toward mathematics scale". Next, the experimental group (39 students) was taught geometry for eight weeks by lesson plans and worksheets designed to improve the students' ability to use metacognitive skills for solving problems. In the meantime, the control group (36 students) was taught through traditional methods. In the end, the three data collection instruments were administered to both groups as a posttest. The data were analyzed via dependent and independent t-tests, and it is seen that the experimental group had significantly higher posttest scores when compared to the control group. More significant results were obtained when the findings revealed by statistical analyses were accompanied by student essays. It was observed that the students in the experimental group had developed a better attitude toward geometry and mathematics, which might be attributed to the improvement in their self-confidence. Furthermore, these students had developed the ability to perceive the importance of problem solving, to understand problems, to be involved in planned studying, and to control and be aware of the problem solving process. The improvement in their attitude toward geometry and mathematics led to a corresponding increase in their achievement.