Surface temperatures for a series of colored specimens were determined under typical exterior exposure and laboratory accelerated test conditions. The temperature separation of colored specimens exposed outdoors depends primarily on the total solar irradiance. Linear regression models can be used to predict surface temperature of colored specimens as a function of the temperature of a black panel. There is good temperature agreement between different types of black panels. In artificial accelerated testing, temperature separations similar to those obtained in typical exterior exposures can only be achieved by appropriate radiant heating from the device's light source. High irradiance conditions in xenon arc devices can result in slower than anticipated failure rates for lighter colored specimens when using black panel temperature control.