Response surface methodology and artificial neural network were used to optimize cultural conditions of L-glutaminase production from Bacillus cereus MTCC 1305. ANN model was superior to RSM model with higher value of coefficient of determination (99.97(ANN) > 97.78(RSM)), predicted distribution coefficient (0.9992(ANN) > 0.896(RSM)) and lower value of absolute average deviation (1.17%(ANN) < 18.47%(RSM)). Optimum cultural conditions predicted by ANN were pH (7.5), fermentation time (40 h), temperature (34 degrees C), inoculum size (2%), inoculum age (10 h) and agitation speed (175 rpm) with a maximum predicted production of L-glutaminase 666.97 U/l which was close to experimental production of L-glutaminase 667.23 U/l at simulated optimum cultural condition. The production of L-glutaminase was enhanced by 1.58-fold after optimization of cultural conditions. Simple kinetic models were developed using Logistic equation for cell growth, Luedeking Piret equation for L-glutaminase production and modified Luedeking Piret equation for glucose utilization indicating that L-glutaminase fermentation is non growth associated process. (C) 2013 Elsevier Ltd. All rights reserved.