In this study, the temperature and viscosity-dependent methods were used to identify the main heat conduction mechanism in nanofluids. Three sets of experiments were conductedto investigate the effects of Brownian motion and aggregation. Image processing approach was used to identify detailed configurations of different nanofluids microstructures. The thermal conductivity of the nanofluids was measured with respect to the dynamic viscosity in the temperature range between 0 and 55 °C. TheResultsclearly indicated thatthe nanoparticle Brownian motion does not play a significant role in heat conduction of nanofluids, which was also supported by the observation that a more viscous sample rendered a higher thermal conductivity. Moreover, the microscopic pictures and the differences in the viscosity between theoretical and experimental valuessuggested the major role of particle aggregation and clustering.