A plethora of work has been done to examine the dynamics of learning in higher education contexts to foster positive learning outcomes. The 3P model (Biggs in Teaching for quality learning at university, Open University Press, Birkshire, 2003) is one theoretical model that has been well researched. However, much of the work to date has focused on examining relationships of two factors (either presage and process; or process and product) of the model, with relatively less effort made to test the interactivity among the three factors in entirety. The present study examined the variables from all three factors of the 3P model longitudinally. Specifically, epistemic beliefs were tracked as the presage variable, alongside students’ perceptions of the learning environment and use of study... approaches as process variables, finally with academic performance and epistemic belief at later time point as product variables. Two hundred and eighty-four Chinese college students participated in the study. Selfreported instruments were used to measure epistemic belief at two time points (Lee and Chan in Asia Pac Educ Res 24(4):603–612, 2015), approaches to learning (Biggs et al. in Br J Educ Psychol 71(1):133–149, 2001) and perception of learning environment (Taylor et al. in Int J Educ Res 27(4):293–302, 1997). A structural equation model showed that epistemic beliefs at Time 1 yielded a direct effect on academic performance without being mediated by any process variables; there seemed to be a disjuncture from presage to product variable as conjectured in the 3P model, and this warrants future investigation.