This study aimed to develop an instrument with acceptable validity and reliability for assessing two key aspects among science teachers: their self-efficacy in teaching science through an integrated STEM approach (STSIS) and their outcome expectancy when employing this approach (OETSIS), which we refer to as TSTSIS. We administered this tool to 109 science teachers in Phnom Penh city and various provinces in Cambodia. The results revealed that STSIS can be broken down into six essential latent factors. Among these factors, five are grouped under the theme of "Integrated STEM Instructional Self-Efficacy," covering problem-based, robots-based, inquiry-based, engineering-based, and technology-based instructional selfefficacy. The sixth factor relates to teachers' self-efficacy in terms of accessing materials. Author formulated the subscale "Expectations of High Students' Achievement in Science" as the factor representing the OETSIS. Scores on the TSTSIS subscales offered evidence of its criterion validity, with significant differences observed across various teacher-related variables. These variables include teachers' teaching qualifications, school types, teachers from schools where STEM education and an integrated STEM approach are part of the school's objectives, and correlation with instructors' teaching experience. The TSTSIS subscale lays the groundwork for designing STEM training programs tailored to the needs of science instructors.