Uncertainty of a hydrological model mainly stems from a lack of understanding and knowledge about the real hydrological process. Input uncertainty and parameter uncertainty are considered to be the two major uncertainty sources of hydrological model. Until now, enormous studies have aimed at calibrating model parameters and estimating model uncertainty. However, these studies mainly ascribe the model output uncertainty to the unknown non-physical parameters. In fact, rainfall, especially of weather radar rainfall, is widely recognized as a main error source. There are seldom studies that aim to explicitly describe model input and parameter uncertainty simultaneously. For this reason, in this study, we investigate the combined effects of radar rainfall uncertainty and parameter uncertainty on the model output. A radar probabilistic quantitative rainfall scheme (Multivariate Distributed Ensemble Generator, MDEG) is integrated with a rainfall-runoff model (Probability Distributed Model, PDM) to calibrate model parameters and estimate the model uncertainty. Finally, the simulated flows, together with their uncertainty bands are compared with the observed flows to evaluate the proposed scheme.