Background: To examine levels of reading self-efficacy in college students and further understand sources that foster reading self-efficacy at the postsecondary level, there is a need for a valid scale that can be used for college student populations. Methods: In Phase 1, 160 items were generated and reviewed by a panel of experts and a group of undergraduate students. Of these, 48 items reached consensus for acceptance and were retained for Phase 2 pilot testing. Pilot results were analysed with principal component analysis to identify the scale structure and remove items not strongly associated with components. Phase 3 finalised the scale and confirmed construct validity using confirmatory factor analysis as well as internal reliability with Cronbach's [alpha]. Phase 4 further assessed construct validity through sensitivity analyses that control for age, gender, education level, ethnicity and grade point average (GPA) to evaluate the potential impact of these covariates on model fit. In addition, Phase 4 assessed the relationships between reading self-efficacy and similar constructs (i.e., academic self-efficacy and reading performance) to evaluate convergent validity of the final scale. Results: Based on Phase 2 principal component analysis, three components were retained--'effort', 'enjoyment', and 'processing'. After removing items with high covariance indicated by the Phase 3 testing, the final scale included 15 items. The Sources of Reading Self-Efficacy Scale for College and University Students (SOURSES-C) was significantly correlated with reading performance, GPA and general academic self-efficacy. Conclusions: SOURSES-C is a valid and reliable tool to describe sources of reading self-efficacy in college students. Understanding of reading self-efficacy profiles may inform further investigations to improve motivation and persistence, self-regulation and strategy use during reading.