Nanofluids, as new heat transfer fluids, are at the center of attention of researchers, while their measured thermal conductivities are more than for conventional heat transfer fluids. Unfortunately, conventional theoretical and empirical models cannot explain the enhancement of the thermal conductivity of nanofluids. Therefore, it is important to understand the fundamental mechanisms as well as the important parameters that influence the heat transfer in nanofluids. Nanofluids' thermal conductivity enhancement consists of four major mechanisms: Brownian motion of the nanoparticle, nanolayer, clustering, and the nature of heat transport in the nanoparticles. Important factors that affect the thermal conductivity modeling of nanofluids are particle volume fraction, temperature, particles size, pH, and the size and property of nanolayer. In this paper, each mechanism is explained and proposed models are critically reviewed. It is concluded that there is a lack of a reliable hybrid model that includes all mechanisms and influenced parameters for thermal conductivity of nanofluids. Furthermore, more work needs to be conducted on the nature of heat transfer in nanofluids. A reliable database and experimental data are also needed on the properties of nanoparticles.