Recent research on achievement in science asserts that motivation, emotion, and metacognition are important driving forces for learning. This study sought to examine the relationships between two physics class emotions (hope and anxiety), their motivational predictors (instrumentality and self-efficacy), and their effects on metacognitive problem solving strategies (planning, monitoring, and evaluation) and performance. Data were collected from 520 grade 11 Spanish students (54.7% girls). Structural equation models (SEM), followed by a bootstrap procedure, were used to examine direct and mediated relationships. The results supported the model, suggesting that instrumentality and self-efficacy negatively predicted anxiety, and enhanced hope, planning, monitoring, evaluation, and performance; metacognitive strategies and performance were negatively predicted by anxiety, and were positively predicted by hope; metacognitive strategies positively predicted performance. Furthermore, the hypothesized mediated relations were also statistically significant. The interpretation of these findings, their implications for physics teaching and learning, and future lines of research are discussed.