This paper describes an improved numerical simulation study of an isothermal chamber recently constructed at Zografou Campus of the National Technical University of Athens (NTUA) for the testing of special equipment used for transporting perishable foodstuffs in accordance with the United Nations ATP agreement. Using a transient finite difference model, a simulation is developed for a modern ATP test chamber and a typical specimen refrigerated vehicle to be tested. The simulation results are compared to experimental measurements taken under real conditions by a data acquisition system and a refrigerated semitrailer as specimen. Proportional-integral control is employed for the regulation of the cooling and heating system. The impact of various parameters on the time required to reach the set-point temperature (is,,) is investigated and the energy consumption is simulated for a period of 22h. In particular, the impact of specimen insulation thickness and the thickness of the chamber insulation floor are considered in detail. The total energy consumption increases by approximately 16% when the concrete floor layer thickness is increased from 8 to 16cm for typical initial conditions and desired chamber and specimen temperatures of 32.5 and 7.5degreesC, respectively. Using a floor insulation of 6cm extruded heavy strain-resistant polystyrene reduces the energy consumption by at least 13%. Specimen insulation thickness increase from U-value of 0.35Wm(-2) K to 0.75 W m(-2) K result to an increase in energy consumption by a percentage of 28%. Thermal capacity, temperature of car body and specimen dimensions are also treated as variables that affect the total duration of an ATP test and its total energy consumption. Copyright (C) 2004 John Wiley Sons, Ltd.