This meta-analysis examines the effects of problem-based learning (PBL) on knowledge and skills, and explores potential moderators of these effects. The analysis includes 43 empirical studies conducted in real-life classrooms, focusing on tertiary education. The results indicate a robust positive effect of PBL on students' skills, with no studies reporting negative outcomes. However, there is a tendency towards negative results when considering the impact of PBL on students' knowledge, though this is influenced by two outlier studies. The combined effect size for knowledge is non-robust. Methodological factors, student expertise level, retention period, and assessment methods are investigated as moderators. The analysis suggests that higher-quality research diminishes the negative effect of PBL on knowledge. Student expertise level is associated with variation in effect sizes, with consistent positive effects for skills. Students in PBL remember more of their acquired knowledge, possibly due to better elaboration and recall. The positive effect of PBL on skills appears immediate and lasting. Assessment methods influence the findings, with more instruments capable of evaluating skills showing larger effects of PBL. The meta-analysis supports the positive impact of PBL on skills and provides insights into moderators, contributing to the understanding of PBL's effectiveness in educational settings.This meta-analysis examines the effects of problem-based learning (PBL) on knowledge and skills, and explores potential moderators of these effects. The analysis includes 43 empirical studies conducted in real-life classrooms, focusing on tertiary education. The results indicate a robust positive effect of PBL on students' skills, with no studies reporting negative outcomes. However, there is a tendency towards negative results when considering the impact of PBL on students' knowledge, though this is influenced by two outlier studies. The combined effect size for knowledge is non-robust. Methodological factors, student expertise level, retention period, and assessment methods are investigated as moderators. The analysis suggests that higher-quality research diminishes the negative effect of PBL on knowledge. Student expertise level is associated with variation in effect sizes, with consistent positive effects for skills. Students in PBL remember more of their acquired knowledge, possibly due to better elaboration and recall. The positive effect of PBL on skills appears immediate and lasting. Assessment methods influence the findings, with more instruments capable of evaluating skills showing larger effects of PBL. The meta-analysis supports the positive impact of PBL on skills and provides insights into moderators, contributing to the understanding of PBL's effectiveness in educational settings.