Precise combustion control and a wider operating load range are the two major challenges in the application of advanced combustion modes, as the combustion process is chemical kinetically driven and thus is sensitive to the fuel composition. Although diesel fuels are complex multicomponent mixtures, most previous research has been carried out using simple single- or two-component surrogate models. This assumption leads to inaccuracies when modeling advanced combustion systems due to differences between the model and real fuel compositions. The present study proposes multicomponent surrogate models for three different diesel fuels that mimic the compositions and property variations of European and American diesel fuels. The composition of the surrogate fuels were arrived at by modeling the distillation data of the three fuels to within 1.5% maximum absolute error. The developed surrogate models were then applied to predict the combustion and emission characteristics of the three fuels tested in a single cylinder diesel engine operated under various conditions, including conventional and low temperature combustion (LTC) conditions. The computations were performed using the multidimensional computational fluid dynamics code, KIVA-ERC, incorporated with suitable reduced reaction mechanisms for the surrogate fuels. The predicted combustion and emission characteristics showed good agreement with measured data. The peak pressures are predicted to within 9.8% mean absolute error. Sensitivities to fuel type and EGR concentration were also explored in conventional and LTC modes using the surrogate models. The results showed that the combustion trends in conventional combustion are less affected by fuel or EGR changes, while a much higher sensitivity was observed under LTC conditions, thus demanding more realistic fuel models to precisely describe advanced combustion modes.