In accelerated light fastness testing the display life environment is simulated by choosing defined constant humidity levels or by cycling between a dry/light and a humid/dark level. The humidity level is a factor that may cause dye diffusion effects on one hand and different degradation speed of the colorants on the other hand. The right choice of the humidity level is important for display life testing and display life predictions.In a group of typical inkjet media/ink combinations some exhibited a strong influence on the humidity cycling conditions whereas others showed little effect. As different colorants have a different sensitivity to humidity their relative ranking change when the test parameters change. This changes the colour balance of faded prints and affects the display life predictions drastically. The first fading colour, as well as the colour shift in image degradation are shown to be a function of the sample humidity during the light exposure.On the basis of light fastness measurements in humidity cycled system over several years the statistical variation for display life of a typical inkjet print has been determined which gives an indication about the experimental errors of a cycled system. A proposal for a dual humidity level test is made.