This study examined (a) students’ reported use of metacognitive knowledge (declarative, procedural, conditional) and metacognitive regulation (planning, monitoring, information management, evaluation) when they are doing school work or homework, and the effect of metacognition on school performance in language and mathematics and (b) the role of hope (agency thinking, pathway thinking) in general self-efficacy, in the impact of general self-efficacy on metacognition, and in the effect of metacognition on school performance. One hundred and sixty-five 5th and 6th grade students (83 boys, 82 girls), randomly selected from 10 state primary schools of various regions of Greece, participated in the study. Data gathered at the second school term of the total three terms. The results revealed that: (a) the reported frequency of use of metacognitive knowledge (mainly, conditional) and metacognitive regulation (mainly, monitoring) was at a moderate extent, (b) hope (predominately, pathway thinking) was a positive formulator of general self-efficacy and of its impact on metacognition, but the influential role of the two constructs differed between and within the components of metacognition, (c) the three sets of predictors had complementary and positive effects on school performance but their relative power in influencing it varied between mathematics and language and within each school subject, with agency thinking being the most powerful predictor and (d) general self-efficacy mediated the impact of metacognition on school performance, while hope had direct impact on school performance beyond that of metacognition and general self-efficacy. The findings are discussed for their practical applications in education and future research.