During 1998, the US Federal authority introduced a requirement for vehicles powered by heavy duty diesel engines that NOx emissions shall be less than 4 g/bhp.h. This represents a 20% reduction over current levels and has prompted significant further hardware changes. As a result of these increasingly tighter NOx emission constraints, soot loading of diesel engine lubricants - due to retarded fuel injection, is becoming an ever more significant issue in crankcase lubricant formulation.For this reason, increased understanding is required of the mechanism of soot particle aggregation and resultant aggregate morphology - together with the likely consequences for the performance of soot-laden lubricants, for viscosity increase, filter blocking, sludging and (directly or indirectly) - soot-induced wear.We describe here a combined experimental and simulation approach to screening formulated lubricants and characterising soot aggregate structures. By using transmission electron microscopy combined with image analysis we can characterise the aggregate structures from soot-laden engine oils (e.g. Mack T-8 and Cummins M-11 tests) in a two dimensional imaging approach. Utilising carbon black-loaded fresh oils, we have extended this to a screening technique.To provide further understanding, we have developed a convenient two dimensional simulation model which uses the essentials of the diffusion limited cluster aggregation (DLCA) approach. This has been modified to allow for variable levels of colloidal stability in the primary soot particles. The simulation model is lattice-based, so computationally efficient and can allow for particle or cluster aggregation under a variety of particle loading regimes and in conjunction with variable levels of particle colloid stability.With the simulation model and experimental data, we can begin to understand such factors as dispersant architecture variables and soot loading rate influences on aggregate morphology and their effects on dispersion viscosity development in the lubricant.