A simplified heat transfer model to predict spatial and temporal temperature variations was developed for insulated boxes equipped with a phase change material (melting point ∼ 0 °C) on a sidewall and loaded with a test product (Tylose). This model, based on the zonal approach, considers the product as four blocks with shell and core regions. It takes into account airflow, heat exchange by conduction, convection and radiation. The model was validated by comparing the predicted temperatures with those measured at 24 positions (on the walls, on the product surface, shell and core, and internal air). The model was first validated under steady state (at 10 °C and 20 °C ambient temperatures), and it showed good agreement with the measured values with a root mean square error of 0.60 °C and 0.88 °C for 10 °C and 20 °C ambient temperatures respectively, and with a regression slope between the numerical and measured curves close to 1 (0.89 and 0.98 for 10 °C and 20 °C ambient temperature, respectively). The model was then validated under transient state (at 4 °C and 10 °C initial product temperatures, 20 °C ambient temperature) and it provided a good prediction of the wall and product core temperature evolution. This model could be a useful tool for stakeholders as it enables them to study the effect of the box configuration e.g. box insulation and operating conditions e.g. the ambient temperature on temperature evolution in real scenarios. By combining this model with a quality and a microbiological model, the quality and safety evolution of a food product could be predicted.