The objective of the study was to test the content validity and construct validity of non-cognitive skills assessment applied to science subjects at Junior High Schools in Indonesia. Non-cognitive skills are traits and skills that can be used to measure non-cognitive factors involved during learning. The study consisted of eight factors namely, critical thinking, open-mindedness, accuracy, perseverance, carefulness, ability to innovate, responsibility, and sensitivity. The content validity method involved 40 experts and panelists (20 teachers and 20 students). The assessment used in this study was self-assessment technique. Exploratory factor analysis was used as construct validity assessment technique and involved as many as 210 students. Viewing the cut-off value of the CFI < 0.96 dan SRMR > 0.06 meant that all eight factors were qualified. The result of the Geomin Rotation Analysis was found valid in 17 items, and 14 items confirmed the cross-loading criteria. The implications of this study included that the instruments’ items do not have to be seen through variables; if the variables are not qualified, the dropped items could be reviewed again through indicators and item editors. Moreover, constructs formed in this study are transferable and replicable, and can be used in future studies, even to test instruments using a large sample and to search item bias. The study recommends using this instrument on teachers and parents too to know the extent of students' non-cognitive skills.