In this study, the effects of water-to-cement ratio (W/C), cement content and coarse aggregate content on the density, void ratio, infiltration rate, and compressive strength of portland cement pervious concrete (PCPC) were investigated by statistical modelling. Two-level factorial design and response surface methodology (RSM) were used. The PCPC mixtures were made with W/C in the range of 0.28-0.40, cement content in the range of 350-415 kg/m3and coarse aggregate content in the range of 1200-1400 kg/m3. In addition, examples were given on using multi parametric optimization to produce isoresponses of a desirability function for PCPC satisfying specified criteria including cost. The results show that W/C, cement content, coarse aggregate content and their interactions are key parameters, which significantly affect the characteristic performance of PCPC. The statistical models developed in this study can facilitate optimizing the mixture proportions of PCPC for target performance by reducing the number of trial batches needed. © 2012 Elsevier Ltd. All rights reserved.