A two-factor central composite rotatable design and corresponding response surface analysis were successfully applied to data from pork and lamb loin roasts, using cooking temperature and endpoint temperature as the two independent variables. Heating rate ("C/min), evaporation loss, cooking time, total moisture, and chromaticity coordinate (z) were found to be significantly affected (p 5 0.10) by different combinations of cooking temperatures and endpoint temperatures in pork and lamb. Additional significant variables for pork were total nitrogen, remaining protein fraction and sensory panel juiciness; and for lamb, heating rate ("C/g), drip loss, total cooking loss, expressible moisture index, total nitrogen, chromaticity coordinate (x), saturation index, and sensory panel doneness and color. Response surfaces were useful in evaluating results.