Background: Problem-Based Learning, PBL, is a student-centered approach in science education that promotes conceptual understanding and higher-order thinking by engaging learners with authentic contextual problems. Aims: This systematic review synthesizes how PBL is implemented in science education, what outcomes are most frequently reported, and what barriers commonly hinder its adoption. Method: A PRISMA-guided systematic literature review was conducted using Scopus. Searches employed the terms “problem-based learning,” “science education,” and “science learning.” Eligible studies were empirical journal articles and conference proceedings published from 2017 to August 2025, written in English, and involving participants from elementary school through higher education. Of 388 records identified, 53 studies met the inclusion criteria and were analyzed according to publication trends, research designs, implementation strategies, targeted outcomes, and reported constraints. Results: Across educational levels, PBL was consistently associated with cognitive gains, especially in critical thinking, problem solving, and higher-order thinking skills. The reviewed evidence also indicated affective benefits, including stronger scientific attitudes, improved learning motivation, and enhanced collaboration. Nevertheless, recurring challenges emerged, such as limited instructional time, demanding assessment processes, curriculum coverage pressures, uneven teacher readiness, unequal participation in group work, and gaps in access to learning resources and technology. Conclusion: PBL offers clear promise for improving both cognitive and affective dimensions of science learning, yet its success depends on well-designed contextual problems, adequate pedagogical support, and alignment with learners’ readiness. Strengthening teacher capacity and refining assessment practices are essential to reduce persistent implementation barriers.