This paper provides a statistical comparison between seismic demands of reinforced concrete (RC) bridges with skew-angled seat-type abutments (or simply 'skewed bridges') subjected to past events using simulations and actual recordings. Three short bridges located in California are selected as seed bridges, from which different models are developed by varying key bridge structural parameters such as column-bent height, symmetry of span arrangement, and abutment skew angle. Through extensive nonlinear dynamic analysis conducted using hybrid broadband simulated ground motions and real records for two historical earthquakes; i.e., the 1989 M 6.8 Loma Prieta earthquake and the 1994 M 6.7 Northridge earthquake, it is demonstrated that the distributions of column drift ratios, deck rotations and displacements produced by simulations agree reasonably well with those produced by recorded ground motions. Statistical hypothesis testing and information theory measures are proposed to quantitatively assess the statistical significance of the results for all the considered demand parameters. Finally, ground motion intensity measures (IMs) related to ground motion directionality and directivity, particularly affecting seismic response of skewed bridges, are 2 compared for both simulations and recorded waveforms. These types of validation exercises can highlight the similarities and differences between simulated and recorded ground motions. The similarities should provide confidence in using the simulation method for bridge engineering applications, while the discrepancies, should help in improving the generation of synthetic records.