Student retention models were tested via structural equation modeling to examine the interrelations and predictability among socioeconomic status, psychosocial, and student success variables with a sample of 445 undergraduate students attending a large Hispanic serving institution. The proposed theoretical model included socioeconomic status (generational status & Pell grant eligibility), psychosocial (academic efficacy, problem solving, connectedness to professors and college), and student success variables (SAT scores, college GPA, intent to remain, and retention). Results provided support for the proposed model and showed that the psychosocial variables examined herein play an important role in predicting connectedness and student success variables. Implications for institutions are discussed.