Modelling growth as a function of feeding rate (FR) could be one of the most important tools for fish farms, because this knowledge allows growth to be maximized, or the feed conversion ratio (FCR) to be minimized, thereby improving profits. All growth models should include the three principal variables involved in growth: initial body weight, temperature and feeding rate. The thermal-unit growth coefficient (TGC) already includes water temperature variation and initial body weight. Studying variation in TGC for fish fed the same diet, but at several feeding rates provides interesting information for modelling. Six different trials were conducted where gilthead sea bream of several different initial weights (24, 38, 50, 110, 220 or 289 g) were fed increasing amounts, and growth and the conversion index response were measured. The TGC response was modelled as a function of FR, and both asymptotic and quadratic responses were examined. The asymptotic model, TGC × 1000 = 2.037*(1−e (−0.8*(FR−0.22)), had an adjusted R2 value of 96.18, whereas the quadratic model, TGC × 1000 = −0.381 + 1,715 × FR−0,382 × FR2, had an adjusted R2 value of 96.42. Simulations of the FCR and the economical profitability index (EPI) were conducted to provide tools for maximizing efficiency and profitability, and the results suggest that these tools will be useful for future investigations.