Abstract: A Bacillus species 2129 was tested for xylanase production under submerged cultivation conditions. Maximum xylanase activities (2.2 U mL-1) were achieved at the following test conditions: substrate concentration-1.7%, pH-6.807 and NH4Cl concentration-0.3%. A simple Artificial Neural Network (ANN) model was formulated using oat concentration, pH and NH4Cl concentrations as the input parameters to the ANN model. The predictions for xylanase activity from the ANN model showed that a multilayer network (3-6-1) using the standard back propagation algorithm was able to predict xylanase activity effectively with R2 values of 0.9995.