In an effort to find correlations between size changes with temperature of lipophilic polymers in solution and viscosity index trends, the determination of the size of thermoresponsive polymers of various architectures (linear, comb-like, star, and hyperbranched) using two experimental techniques under infinite dilution conditions (0.5% w/w) - dynamic light scattering and small angle neutron scattering, and predictive molecular dynamics simulations is described herein. Viscosity index is an important parameter for lubricants and other rheological applications. The aim of this work was to predict polymer behavior as viscosity index improvers (VIIs) using tools which require minimal amounts of material, as opposed to measuring kinematic viscosities, which require multigram quantities. There were no significant correlations between changes in polymer size with temperature and viscosity index (VI). The polymers with the highest VI (polyalkyl methacrylate - PAMA and Star PAMA) had polar backbones in contrast to the nonpolar backbones of the and hyperbranched (OCP and HBPE, respectively), so the disparity in solubility of the backbone and solvent medium appears to correlate with the observed VIs. It was concluded that none of the aforementioned techniques can entirely predict the polymer behavior as VIIs, at least in the temperature range studied (40-100 degrees C).