Bayesian stochastic inversion (BSI) and Multi-Criteria (MC) methodologies were applied to the dataset from Cabauw, Netherlands, to determine the optimal parameters and uncertainty for the CHASM (CHAmeleon Surface Model) land surface model. Simulations were performed to calibrate CHASM against observed sensible heat fluxes, latent heat fluxes and ground heat fluxes. The calibration resulted in very similar 'best fit' models betwwen the BSI and MC methodologies that show better performances than the default model, although the MC method did not give a consistent view of the true uncertainty. Through evaluations of model performance that considered the seasonally averaged diurnal cycle and monthly averages of energy fluxes, the three most important CHASM parameters are minimum stomatal resistance, vegetation roughness length, and vegetation fraction cover. One of the strengths of the BSI methodology is that it can express uncertainty in terms of probability density functions and error covariances that indicate non-linear relationships between model parameters. With these estimates of uncertainty we can identify which sets of model parameters would yield predictions that are within published error estimates in the Cabauw data and provide an objective way to investigate how uncertainty in the CHASM model would affect GCM model predictions of future climate change. Although we have made no effort to optimize the BSI algorithm in the current study, it required about 5 times more computations than the MC algorithm. This is still a substantial saving over the Gibbs sampler that requires at least 10 times more computations than the BSI algorithm to obtain similar results.