Friction stir welding generates significant temperature increases, leading to microstructural changes that influence the mechanical properties of the material. Temperature control is therefore essential to ensure the quality of the welded joint. This study aims to model and predict the maximum temperature generated during the friction stir welding of aluminum alloy 3003, based on three key operating parameters: rotation speed, feed rate, and tool inclination angle. The response surface method was used to develop a robust predictive model and evaluate the individual and combined effects of these parameters on the thermal response. The results reveal that the most influential parameters are, in order, rotation speed, tool inclination angle, and feed rate. They also indicate that the maximum temperature increases significantly with rotational speed and angle of inclination. In contrast, it decreases as the feed rate increases. The model obtained has excellent predictive power, validated by a low root mean square error of 4.41 °C and a coefficient of determination R² of 0.972.