To determine the influence of nutrients on the rate of biodegradation, a five-level, three-factor central composite design (CCD) was employed for bioremediation of seawater artificially contaminated with crude oil. Removal of total petroleum hydrocarbons (TPH) was the dependent variable. Samples were extracted and analyzed according to US-EPA protocols. A significant (R(2) = 0.9645, P < 0.0001) quadratic polynomial mathematical model was generated. Removal from samples not subjected to optimization and removal by natural attenuation were 53.3% and 22.6%, respectively. Numerical optimization was carried out based on desirability functions for maximum TPH removal. For an initial crude oil concentration of 1 g/L supplemented with 190.21 mg/L nitrogen and 12.71 mg/L phosphorus, the Design-Expert software predicted 60.9% hydrocarbon removal; 58.6% removal was observed in a 28-day experiment. (C) 2010 Elsevier Ltd. All rights reserved.