The world is experiencing more intense and frequent extreme precipitation events due to climate change. At the weather time scale, recent studies based on observational and reanalysis data found that extreme precipitation increases with near-surface air temperature up to a threshold and then decreases as temperature further increase; however, extreme precipitation increases monotonously with near surface air temperature if saturation deficit is kept the same, and the rate of increase does not deviate too far from the Clausius-Claperon ratio. It is not clear how different earth system models perform in capturing such relationships and how model resolution may impact the performance. DOE’s Energy Exascale Earth System Model (E3SM) is a state-of-the-science model that has been developed and tested for different spatial resolutions. In this study, we aim to assess how E3SM performs in reproducing extreme precipitation and the emergent relationship between extreme precipitation and temperature, and examine how finer spatial resolution may improve the model performance. We performed both univariate analyses on how extreme precipitation (e.g., 99^th percentile) scales with near-surface air temperature and bivariate analyses on how extreme precipitation scales with near-surface air temperature and saturation deficit or dew point depression. Results from the analysis showed that the scaling relationships from the E3SM model are qualitatively similar to those from the observational and reanalysis data, but some differences exist in the temperature at which extreme precipitation peaks and the rate of extreme precipitation scaling with temperature in a saturated atmosphere. To assess how the model resolution may influence its performance in reproducing the extreme precipitation-temperature scaling, we analyzed output from E3SM runs at two different resolutions, 25km and 100km, This presentation will compare results from the two different resolutions and quantify the impact of spatial resolution changes on model performance in reproducing the observed scaling relationships.