The CHAmeleon Surface Model (CHASM) is used with the Project for the Intercomparison of Land-surface Parameterization Schemes (PILPS) Phase 2(e) experimental design. The model, with five levels of surface energy complexity, is shown to simulate the observed runoff for two sub-basins within the PILPS 2e domain with some degree of skill. This skill is shown to be largely insensitive to the complexity of the surface energy balance (SEB) parameterization, provided that a spatially and temporally constant surface resistance is used. While all bar the least complex mode of CHASM works reasonably (within 5% of the observed runoff at Ovre Abiskojokk and within 15–20% at Ovre Lansjarv), there are weaknesses in the simulations, particularly at a seasonal time scale. It is shown that these deficiencies are unlikely be caused by the complexity of the surface energy balance formulation. At the scale of the PILPS 2(e) domain, it is shown that the addition of a constant surface resistance into the simplest mode of CHASM has a large impact on runoff by reducing the latent heat flux. It is shown that adding further complexity by including explicit interception and explicit bare soil evaporation has little effect on runoff or the latent heat flux, but the addition of a spatially and temporally variable surface resistance leads to an increase in the spatial variability of runoff across the domain. The differences between the modes of CHASM are then compared against the range of simulations of other models in PILPS Phase 2(e). The difference between the modes of CHASM (excluding the simplest mode, which omits the surface resistance term) is approximately 10–15% of the range in results from the other PILPS schemes. This is argued to be too small to explain the differences found in the results from PILPS Phase 2(e). Given that differences cannot be explained by variations in the SEB complexity, they are most likely to be caused by differences in the hydrology formulation (including snowmelt and snowmelt infiltration).