A classic H 2 SO 4 -H 2 O binary homogeneous nucleation model coupled to an aerosol dynamics model, suitable for studying the formation and transformation of volatile nanoparticles (VNPs) during diesel engine exhaust dilution, has been developed. Using the H 2 SO 4 -H 2 O binary homogeneous nucleation model, the nucleation ratio and molecular cluster size were calculated. The effect of aerosol dynamic processes on VNP number size distributions was studied. The effects of fuel sulfur content (FSC) and sampling conditions in the laboratory on VNP number size distributions were also calculated. Our simulations demonstrated that nucleation increased the cluster number concentration and that FSC, temperature and humidity significantly affected the nucleation ratio and molecular cluster size. Coagulation promoted the evolution of cluster-particle size distributions from monodisperse to polydisperse. Soot present in the exhaust can suppress the formation of VNPs. FSC and sampling conditions, like the primary dilution temperature, the primary dilution relative humidity, residence time and the primary dilution ratio have significant effects on VNP number size distributions. . VNP number size distributions are nonlinear and highly sensitive to the temperature, humidity and dilution ratio during the dilution process. Recently, theoretical studies on VNPs in diesel engine exhaust have been conducted [7,8]. These studies largely focused on the evolution of VNPs in the atmosphere, leaving many questions about the mechanisms driving the formation and transformation of VNPs. Further research in this area is necessary. In this study, a classic H 2 SO 4 -H 2 O binary homogeneous nucleation model coupled to an aerosol dynamics model is presented. The work in this paper includes: (1) the nucleation ratio and molecular cluster size based on binary H 2 SO 4 -H 2 O homogeneous nucleation theory, (2) the effect of the aerosol dynamic processes on VNP number size distributions and (3) the effects of fuel sulfur content (FSC) and sampling conditions in the laboratory on VNP number size distributions. The experimental results were then compared with the modeling results.