The present study envisages the performance of a laboratory scale electrocoagulation system for the removal of Cr(VI) from 100 mg l(-1) solution using Al-Al electrodes with an effective surface area of 100 cm(2), and placed 15 mm apart. The interaction between voltage x time, and amperage x time best explained the Cr(VI) reduction efficiency with the coefficient of determination (R-2) being 0.8873 and 0.9270 respectively. Similarly, the square root of energy consumption in Cr(VI) reduction had a linear correlation with voltage x time (R-2 = 0.8949), whereas. amperage x time better explained energy consumption (R-2 = 0.9400). Response surface methodology was used for the optimization of process variables (pH, voltage and treatment time), response modeling and predictions. Maximum Cr(VI) reduction efficiency of 90.4% was achieved at pH 5, 24 V and 24 min treatment time, and the treatment consumed 137.2 KWh m(-3) of electrical energy. Multiple response optimization for maximizing Cr(VI) reduction efficiency and minimizing energy consumption showed 49.6% Cr(VI) removal at pH 5, 12.8 V and 24 min treatment time. The response models developed explained 95.2% variability for Cr(VI) reduction efficiency and 99.4% variability for energy consumption. Results of the prediction models were validated through laboratory scale batch experiments. (C) 2009 Elsevier B.V. All rights reserved.