Nanofluids are engineered suspensions of fine nanoparticles in basefluids. Owing to their enhanced thermal conductivity, nanofluids find applications in many heat transfer and energy conversion systems. Enhanced thermal conductivity of nanofluids is attributed to several mechanisms such as Brownian motion, interfacial layer formation and particle clustering, etc. Many theoretical models have been proposed based on these mechanisms to predict the thermal conductivity of nanofluids. But, still there is an uncertainty in predicting the thermal conductivity of nanofluids. In this work, a simple model to predict the thermal conductivity of nanofluids based on particle size distribution and multi-level homogenization has been proposed. This model considers the effects of Brownian motion, interfacial layer formation and particle clustering. Particle clusters are characterized based on particle size distribution (PSD) analysis and their thermal conductivity is calculated exclusively. The complex nanofluid system is subdivided into smaller systems and a level by level homogenization is carried out to determine the effective thermal conductivity of nanofluids. Present model predictions are compared with experimental results from literature and are found to match well. Contributions of aggregation, Brownian motion and interfacial layer formation are individually exhibited. This model aids to develop a better understanding of the thermal transport in nanofluids and hence, is expected to contribute to several industrial applications. (C) 2016 Elsevier B.V. All rights reserved.