There are situations where teachers accustomed to years of traditional teaching are tasked with implementing inquiry-based science teaching in schools nationwide, with few professional developments to change their self-efficacy. The implicit expectations are that teachers' self-efficacy will improve as they implement inquiry-based practices over time. This study explored self-efficacy levels of science instructors outside of the professional development context; and combined concepts from social cognitive and curriculum implementation theories and findings to explore effective ways of implementing inquiry-based teaching practices in schools. We collected, analysed, and triangulated quantitative and qualitative data from a larger sample of 308 and subsample of 18 junior high school science instructors. We found that science instructors outside the professional development context had low self-efficacy of inquiry-based but high self-efficacy of traditional teaching practices. We also found that there were statistically substantial differences in self-efficacy of science instructors outside of the professional development context, based on school type, school location, and academic qualification. From combination of concepts from social cognitive and curriculum implementation theories and findings from the study, it was ascertained that differentiated and phased approach to implementation of inquiry-based teaching practices needs to be adopted in various schools of diverse teachers with different self-efficacy levels. Different teachers in different schools with different levels of self-efficacy may start implementing inquiry-based practices with different content knowledge, teaching skills, and progress at different rates.