The purposes of this study were: 1) to develop the measurement instruments of junior high school students´ metacognition, motivational belief, and mathematics problem-solving strategies; 2) to investigate the relationship between junior high school students´ metacognition, motivational belief, and mathematics problem-solving strategies; and 3) to compare the differences of mathematics problem-solving strategies between junior high school students with different metacognition and motivational belief. There were sixty subjects (male 30, female 30) drawn from two junior rhigh schools (Chang Ho and Tian-Mu) in Taipei area. All the subjects were administered by Metacognition Inventory, Motivational Belief Inventory, Mathematics Test, and Mathematics Problem-solving Process Inventory. The obtained data were analyzed with Pearson product-moment correlation, canonical correlation, and t-test. The main findings of this study were as follows: 1. There were relationships between junior high school students´ metacognition, motivational belief, and mathematics problem-solving strategies. 2. Junior high school students´ metacognition and motivational belief could account for 40% of the total variance in their mathematics problem-solving strategies. 3. There were significant differences in mathematics problem-solving strategies between the junior high school students with different metacognition and motivational belief.