In this study, a latent growth curve model of academic motivation and academic self-efficacy in Shahid Chamran University students was tested. The research method was developmental and of a longitudinal type. More precisely, the latent growth curve model was used to answer research questions. The statistical population of this study was all newly-arrived freshmen students of Shahid Chamran University in Ahvaz in the academic year of 2016-2017, who were 1988 students. Random sampling method was used to select newly-arrived students. From 13 departments of Shahid Chamran University of Ahvaz, 4 were randomly selected and all of their freshmen students filled out the questionnaires. The sample size of this study was 678 students. The instruments of this study include Academic Motivation Scale (AMS) and Academic Self-Efficacy Scale. Results showed that the mean score of intercept of academic motivation (at the beginning of university studies) was higher than the mean score of intercept of academic self-efficacy. Although the mean score of slopes of academic motivation and self-efficacy were decreased among the students during the three measurement times in university, the self-efficacy slope was not significant. Also, students who had high academic motivation at the beginning, had high scores in academic self-efficacy too, and those who had made progress in their academic motivation during the one year study, have improved their academic self-efficacy too. Moreover, all the variances were significant. The results showed that the interpersonal variations support adding the predictor variables to the model. Also, the results indicated that gender was a significant predictor for academic motivation at the beginning of university studies. The results showed that the females' academic motivation, at the beginning of university, on the average, was more than the males’, but the difference was not significant during the first year in university. Also, at the beginning, the academic performance was a significant predictor of academic self-efficacy. In other words, at the start of university studies, on the average, the high academic self-efficacy was associated with high academic performance, but the difference was not significant during the three measurements. Moreover, the non-linear model's goodness of fit statistic was better than the linear models.