Climate-based daylight modelling is based on the available weather data, which means that the weather data used as input to the daylight simulations are of great importance. In this paper, the effect on the outcome of the daylight simulations of using one weather data file rather than another for the same location was investigated. Furthermore, the effect of using weather data sets with an hourly resolution compared to a one-minute resolution was investigated. The results showed that the lighting dependencies varied by up to 2% depending on the chosen weather data file and indoor illuminance threshold. The energy consumption for artificial lighting was underestimated when simulating with time steps of hourly means compared to one-minute resolution. The findings from this comparison show that the dynamic, short-term effects of the weather have a surprisingly small impact on the simulation outcome.