<span lang="EN-US">To ensure the success of science learning oriented to higher order thinking skills, especially scientific communication skills (SCS), valid and effective instruments are needed. The existing instruments, although helpful, are still too general and have not touched the socio-cultural context. Therefore, we developed an instrument to measure SCS according to the characteristics of students in Indonesia. This study tested the validity, reliability and assumptions of the hypothetical model of the scientific communication skills inventory (SCSI) instrument developed. This study involved 1,287 students in Indonesia. The test was carried out through confirmatory factor analysis with structural equation modelling (SEM) using AMOS version 21. The results of the analysis show that the reliability of the instrument based on Cronbach’s alpha coefficient is 0.720. The convergent validity test detected one item of the oral presentation indicator as invalid and the discriminant validity test confirmed that all instrument items were declared valid. The goodness of fit test shows all scores according to the recommended standard criteria so that SCSI can be used to identify the SCS ability profile of students in Indonesia. For further research, we recommend the development of a learning model that prioritizes the context of social issues to grow students’ SCS.</span>