The aim of this study is to find out whether the science self-efficacy scale in PISA 2006 and PISA 2015 applications ensure measurement invariance. Sample of the study consists of 4791 students in PISA 2006 and 5071 students in PISA 2015 implementation. Multi-group Confirmatory Factor Analysis (MGCFI) was performed to determine invariance of the science self-efficacy scale across year and gender. Invariance stages were examined by means of the comparison of fit indexes. The results of the study indicated that the science self-efficacy scale satisfied all stages of invariance by gender both in 2006 and 2015. Structural and metric invariance was provided for both gender across years and total group across years.