This study aims to look at the effect of Problem Based Learning (PBL) Strategy and mathematical anxiety towards mathematics learning outcomes. The method used is the quasi-experiment method with a 2x2 factorial pattern. The data analysis technique is was the two-way analysis of variance (ANAVA) technique. This study discovers that (1) there are differences in mathematics learning outcomes between students who got the Problem Based Learning (PBL) strategy treatment and conventional learning strategies treatment, (2) there are effects of interaction between learning strategies and anxiety on mathematics learning outcomes, (3) there are differences in mathematics learning outcomes of students who have high anxiety level based on the treatments they received, (4) there are differences in mathematics learning outcomes of students who have low anxiety level based on the treatments they received.