This study uses a domestic dwelling as the setting to investigate and explore the applicability of daylighting metrics for residential buildings, including the formulation of metrics for nonvisual effects. The simulation approach used to generate the performance data from which the metrics are derived is called climate-based daylight modelling (CBDM). This approach delivers predictions of various luminous quantities using sun and sky conditions that are derived from standardised annual meteorological datasets. Although there are uncertainties regarding the precise calibration, there is now sufficient empirical data to parameterise models that also simulate the non-visual aspects of daylight, e.g. for circadian entrainment and a general sense of "alertness". For these non-visual aspects, vertical illuminance at the eye was predicted using a modified climate-based daylight modelling approach. In the paper, we consider what relation there might be between the three aspects of daylight provision and if these relations appear to be complementary or conflicting in nature: for task; to reduce electric lighting usage; and, for non-visual effects. The implications for future building guidelines for daylighting are also discussed.