The interactions of hydrological variables in the terrestrial hydrological cycle are complex. To better predict the variables, distributed and physically based models are used as they account for the complexity of interactions. In this study, we addressed the joint simulation of water- and energy fluxes and the potential benefit of flux measurements in the parameter estimation process. For this purpose, we applied the hydrological model GEOtop to a prealpine catchment in southern Germany (River Rott, 55 km ² ) over two recent summer episodes, as a test case. Due to its complexity, the model is computationally demanding and only a limited number of forward runs can be afforded in inverse modelling and parameter estimation. We applied the gradient-based nonlinear Gauss-Marquardt-Levenberg (GML) parameter estimation method and linked the GEOtop model to the Parameter ESTimation tool (PEST). Using this developed GEOtop-PEST interface, we particularly investigated the value added by including turbulent flux data in the parameter estimation process, and analyse the impact of the additional flux data on the uncertainty bounds of the parameters. To better understand the interplay of the model parameters and to identify the dominating parameters in the calibration process, we also conducted a Principal Component Analysis (PCA). We were able to identify a set of model parameters that reproduced both observed streamflow and turbulent heat fluxes reasonably well. The majority of the estimated parameters were highly sensitive to the considered variables. We showed that the confidence bounds of estimated parameters are narrowed significantly when considering not only streamflow observations but also turbulent flux measurements in the calibration process. In this manner, correlations between estimated parameters could also be reduced.