The aquaculture sector has been steadily growing and thus there is an increasing need to develop mathematical models that allow the estimation of production-related parameters. Prediction of fish growth, feed requirements and waste outputs, are essential in order to ensure the profitability and sustainability of the production activities. Bioenergetic models have been widely used to estimate growth based on energy budgets, but they have some limitations by not explicitly considering the mass-balance of the main macronutrients (e.g., protein). In turn, nutrient-based models are more explanatory, as they consider both energy and nutrient inputs, and estimate fish growth by simulating nutrient accumulation in the fish body. Although some bioenergetic and nutrient-base models for predicting Nile tilapia growth exist in the literature, their suitability is not entirely clear, since their development is often based on uncertain or suboptimal criteria (e.g., relying solely on calibration goodness-of-fit measures).In this work, Nile tilapia growth datasets covering a wide range of rearing conditions and feed compositions were collected from the scientific literature. An exploratory analysis of the collected data was performed to clarify the relationships between energy/protein intake and gain. In this analysis, a direct relationship was observed between digestible energy intake and energy gain, as well as between digestible protein intake and protein gain. Protein gain showed better efficiency than energy gain, even at higher intake levels and without clear evidence of a saturation effect. While digestible energy intake negatively affects energy retention efficiency, digestible protein intake does not significantly impact protein retention efficiency. Furthermore, while energy retention efficiency varies with fish body weight, the same effect was not observed for protein retention efficiency. Finally, though DP/DE ratio has no apparent effect on energy retention efficiency, it seems to negatively affect protein retention efficiency. Considering these observations, plausible growth models with different levels of complexity were developed and calibrated under a diverse set of assumptions. Additionally, two growth models already published for Nile tilapia were calibrated using the same datasets, and their performance was compared with the models developed in the present study. The exploratory analysis of the data suggests a direct relationship between digestible energy intake and energy gain, and between digestible protein intake and protein gain. Furthermore, fish body weight seems to affect the energy retention efficiency. The results of model evaluation showed that energy-protein flux models (EP models) have lower errors in predicting fish growth than pure bioenergetic models (MAPEbw ∼ 9% against ∼13%, respectively), showing the importance of considering protein intake when estimating Nile tilapia growth. Furthermore, assuming the fixed standard metabolic body weight exponents of 0.80 and 0.70 (for energy and protein, respectively), rather than estimating them from the data, seemingly improved the predictive ability of the models. This approach demonstrates the benefits of coupling bioenergetics with nutrient-based models to predict growth and body composition of Nile tilapia along time.