In this study, influences of seven process variables such as initial pH (pH(i)), applied current (i), operating time (t(EC)), initial As(III) concentration (C-o), diameter of Fe ball anode (d(p)), column height in the electrocoagulation (EC) reactor (h) and airflow rate (Q(air)) for removal of As(III) from groundwater by a new air-fed fixed-bed EC reactor were evaluated with a response surface methodology (RSM). The proposed quadratic model fitted very well with the experimental data for the responses. The removal efficiencies and operating costs were determined to be 99% and 0.01 $/m(3) at the optimum operating conditions (a pH(i) of 8.5, 0.05 A, 4.94 min, d(p) of 9.24 mm, h of 7.49 cm, Q(air) of 9.98 L/min for 50 mu g/L). This study clearly showed that the RSM in the EC process was a very suitable method to optimize the operating conditions at the target value of effluent As(III) concentration (10 mu g/L) while keeping the operating cost to minimal and maximize the removal efficiency.