Multiple regression analysis was used to develop mathematical models applicable to the growth and food intake of Octopus vulgaris. The variables considered were: body weight (Bw: 175–3,500 g), temperature (T: 13–28 °C), sex (S: male = 0, female = 1) and diet (D: bogue fish = 0, crabs = 1). Growth and food intake may be succesfully expressed by means of the following equations: Ln (AGR + 14) = −2.0135 + 0.0895 Ln Bw + 0.5087 T − 0.0142 T2 + 0.2997 D (R2 = 71.79; ANOVA p < 0.0001) and Ln (AFR) = − 5.6577 + 0.5137 Ln Bw + 0.5266 T − 0.0132 T2 + 1.1135 D (R2 = 78.71; ANOVA p < 0.0001), where AGR: absolute growth rate, AFR: absolute feeding rate, Bw: body weight, T: temperature and D: diet. In our experimental conditions, sex did not affect growth or food intake. The optimum temperature for growth (17.5 °C) and food intake (20 °C) was independent of body weight. Growth and food intake were higher with the crab diet. Nevertheless, food efficiency was better for animals fed on fish (bogue). Maximum food efficiency was reached at 16.5 °C for both diets. When the temperature was above 23 °C, weight losses and mortality were recorded; the temperature at which this occurred depending on body weight and diet, so that smaller and bogue-fed individuals were more sensitive to increasing temperatures. O. vulgaris growth may provide optimum economic performance from 16 to 21 °C. This range is too narrow, considering the wide natural range (12–29 °C) in some Mediterranean areas. Therefore, O. vulgaris growth will be limited by seasonality of temperature or must be carried out with other systems (e.g. recirculation in closed systems with temperature control) for it to be economically viable.