Viscosity modifiers (VMs) are high molecular weight polymers whose functionality is derived from their thickening efficiency, viscosity–temperature relationship, and shear stability. There are now many different additive chemistries and architectures available, all of which have advantages and disadvantages, and affect solution viscosity through different mechanisms. Understanding these mechanisms and how they impart additive function is critical to the development of new viscosity modifiers that enable lubricants to function more efficiently over a wide range of temperatures. The goal of this dissertation is to investigate the mechanisms underlying commonly used VMs using molecular dynamics (MD) simulations. We first investigated the effect of linear styrene–butadiene polymer structure on the temperature–viscosity behavior of model polymer-base oil solutions. Three different styrene-butadiene polymer configurations were constructed with alternating, random, and block monomer order in a dodecane solvent. Mechanisms underlying this polymer's function were explored by quantifying the coil size and intramolecular interactions. The results indicated that the block styrene-butadiene configuration enabled the formation of small coils with more intramolecular interactions, resulting in the least change in viscosity with temperature, thereby demonstrating a characteristic of a good VM. Next, we studied the effect of polyisobutylene (PIB) on the viscosity of a polyalphaolefin base oil. The mechanisms by which the PIB increased viscosity were explored using the simulations. The results showed that coil size and polymer association did not contribute to solution viscosity. While the results showed that coil size and polymer association did not contribute to solution viscosity, measurements of solvent orientation close to the polymer additive suggested that modification of the solvent by the polymer contributes to viscosity enhancement. Lastly, we investigated hydrocarbon and polymethacrylate comb polymer configurations where the hydrocarbon arms were in different locations on the polymer backbone, and found that association mechanisms may be the underlying mechanism for the observed differences in solution viscosity and performance. In general, the need for optimized VMs will continue to become more important as lubricants are asked to provide better performance under a wider range of operating conditions. This work uses a series of new simulation tools that can study the polymer properties of commonly used VMs through molecular dynamics simulations. Characterizing VMs on the molecular scale is an approach that can provide an understanding of how VMs function and ultimately enable the design of optimized VM polymers.