The aim of the study is to examine the measurement invariance of the science literacy self-efficacy items in the PISA 2015 student questionnaire in different science literacy achievement groups. In this context, the sample of the study consists of 5556 students from Korea representing the high achievement group, 6648 students from Spain representing the medium achievement group, and 5862 students from Turkey representing the low achievement group who responded to the Science self-efficacy questionnaire in PISA 2015. The self-efficacy model for science literacy, which was created by exploratory factor analysis in the first step with 18066 student data from three countries within the scope of the research, was confirmed with 8 items and a single factor. In the second step, it was examined whether the model provided measurement invariance in different success groups. In the process of examining the invariance, multi-group confirmatory factor analysis was performed, and the invariance of the model was examined gradually. The results of the research show that the self-efficacy model for science literacy provides shape invariance, metric invariance and scale invariance in different achievement groups. Since the model provides partial invariance between countries in different achievement groups, comparing the scores of individuals in the countries from the observed variables included in the model will give accurate results.