A nanofluid filled direct absorption solar collector (DASC) system in which incident sunlight is absorbed directly by a working fluid, provides a promising alternative to conventional solar collectors. Most of the previous numerical and experimental studies evaluated the effect of various nanofluids on the thermal performance of a pre-designed collector, and did not consider the effect of varying collector dimensions on its overall performance. In this study, a numerical model of nanofluid flow and temperature distribution in a DASC is proposed by solving the radiative transfer equations of particulate media and combining conduction and convection heat transfer equations. Response surface methodology (RSM) was then applied to understand the effect of varying dimensions on thermal efficiency and entropy generation of the DASC collector design. Based on the produced response surfaces, multi response optimization was performed to find the collector optimized geometry within the studied range of dimensions. (C) 2015 Elsevier Ltd. All rights reserved.