The present study employs social cognitive theory (SCT) and social capital as the guiding frameworks to explain online interaction quality among learners in a blended learning program (N=179). Capturing performance expectancy by perceived learning benefits and online interaction quality with nuanced cognitive measures, the study aims to validate how the SCT and social capital, which are mostly used in predicting online interaction quality in virtual settings, are applicable in an academic setting. More specifically, we investigate the relationship between trust, norms of reciprocity, sense of belonging, altruism, perceived learning benefits and learners’ perception of online interaction quality. Data were collected quantitatively by means of a questionnaire. Confirmatory factor analysis (CFA) and path analysis were employed to validate the instrument and answer the research question respectively. The findings show that only sense of belonging and perceived learning benefits significantly account for a moderate variance in online interaction quality (R2 = .46). Based on the findings, implications for instructional practice and further research are suggested.