Just a few investigations have been conducted to study the mixed nanofluids(MNs), which contain more than one type of nanoparticles, despite considerable advances in the field of nanofluids thermal conductivity. In present research, by combining different volume fractions of various nanoparticles, the variation of mixed nanofluids thermal conductivity was considered. The mentioned nanofluids have different fabrication cost. First, the effect of each specific nanoparticle presence in the base fluid on the thermal conductivity of nanofluid was surveyed both experimentally and theoretically. Then, the thermal conductivities of two MNs, one consisted of a metallic nanoparticle (high thermal conductivity) and an oxide nanoparticle (low thermal conductivity), and the other included two different types of oxide nanoparticles, were measured. Also, a theoretical model based on the Brownian motion, considering the Electrical Double Layer (EDL), was proposed, in order to estimate the enhancement of thermal conductivity of these nanofluids. Contrary to what was expected, the preliminary results indicated reduction of thermal conductivity of MNs, compared to the one containing a single type of nanoparticle; due to nanoparticles agglomeration. The mentioned issue was solved via controlling the pH of nanofluid and using appropriate surfactant. Application of this trend led to minimizing the difference between experimental and developed theoretical results of thermal conductivity of MNs. Generally, anomalous anticipated enhancement of MN's thermal conductivity in comparison to conventional nanofluid was not observed. (C) 2013 Elsevier Masson SAS. All rights reserved.