The purpose of this study was to investigate the effect of using metacognitive strategy training on mathematical problem solving achievement. The study took place over a nine-week period with 47 fifth grade students. The experimental group (n = 24) instructed to improve their metacognitive skills. At the same time the students in the control group (n = 23) received no additional activities and continued their normal lessons. Students were pre- and post-tested with the Mathematical Problem Solving Achievement Test and Turkish version of Metacognitive Skills and Knowledge Assessment (MSA-TR). The results indicated that students in the metacognitive treatment group significantly improved in both mathematical problem solving achievement and metacognitive skills. (Contains 6 tables.)