This study aims to describe and analyse the effect size of the Project Based Learning (PjBL) learning model on the physics learning outcomes of students in high school/ vocational equivalent, calculated based on (1) sample size; (2) duration of the study; (3) subject matter, and (4) a significant influence on the overall PjBL learning model. This type of research is a meta-analysis which is a form of research using pre-existing data. The research population is a Project-based learning model journal with physics learning outcome variables. Hence, research samples in the form of scientific articles published through google scholar nationally are at the high school/ vocational level in 2015–2022 and use experimental research types with predetermined criteria. The research instrument uses a coding sheet. The result of research from 18 articles showed the influence of PjBL on physics learning outcomes, (1) based on the sample size obtained a sample of 72 people with a value of 1,678 very large categories; (2) based on the duration of the study, it was found that the research for five weeks had a value of 1,648 very large categories; (3) based on the subject matter has the result that the global warming material with a value 1,678 categories is very large; (4) the overall effect size show that the PjBL model affect physics learning outcomes with a medium category of 0.45. This study shows that the project-based learning model can be an alternative to improve the physics learning outcomes of students in high school/ vocational school. However, the effect size in the study varies. Therefore, this research implies that to develop learning to improve student learning outcomes, the project-based learning model can be used to further improvements.