Atmospheric cold and warm fronts are a major driver of extreme precipitation over Europe. To assess future changes in extreme weather, it is therefore essential to understand how frontal systems respond to a warming climate. This requires the analysis of climate model projections. A crucial first step is a process-based evaluation of frontal dynamics in present-day simulations, as this increases confidence in the models and the reliability of their future projections.</p> <p>In this study, we compare the representation of frontal frequencies, frontal extreme precipitation, and frontal structure in the CMIP6 and EURO-CORDEX ensembles, using ERA5 as a reference. To assess the added value of higher resolution, we analyze the models on their native grids and compare them with ERA5 data remapped to similar resolutions.</p> <p>We found that all models exhibit substantial biases in frontal frequencies and associated extreme precipitation, which are possibly related to storm-track position biases and an underrepresentation of land–atmosphere interactions. Warm frontal extremes are generally better captured than cold frontal extremes. Increasing model resolution leads to significant improvements for cold frontal biases, whereas warm frontal biases remain largely unaffected. The analysis of frontal structures supports this interpretation: while synoptic-scale conditions are well represented across models, meso-scale gradients and circulation patterns exhibit a pronounced sensitivity to grid spacing. Because warm fronts extend over larger spatial scales, they are already reasonably well simulated at coarse resolution. Cold fronts, by contrast, are governed by smaller-scale processes and therefore show notable improvements at higher resolution.</p> <p>These findings provide an important step toward evaluating climate models in their ability to simulate extreme weather phenomena. While warm frontal extremes appear robust across model resolutions, reliable simulations of cold frontal extremes require higher-resolution models to adequately capture their dynamics and associated extreme precipitation.