The discovery of a novel type of photoreceptor that mediates non-visual light responses in humans has sparked a growing interest in the role of lighting design on human health and wellbeing. Researchers have identified intensity, spectrum, duration/pattern, history, and timing of light exposure as important variables that control the responsiveness of the non- visual system. All of these variables need to be considered when developing a model of non- visual light responses. Currently, there is no mathematical model that incorporates all five variables to predict the non-visual effects of light on humans. In this paper, a modular model structure is proposed towards this end. The model is represented by a sequence of different blocks or elements. Based on a part of this model, which takes into account the intensity, spectrum, and duration of light exposure, it is possible to compare the spectra of different light sources in terms of non-visual driven efficiency. This model provides a framework that can inform designers about how lighting improves human health.