The fate of organic chemicals in the environment depends on their susceptibility to biodegradation. Hence, development of regulations concerning their manufacture and use requires information on the extent and rate of biodegradation. In the paper, a nonlinear group contribution method has been developed using neural networks; it is trained using literature data on the first-order biodegradation kinetic rate constant for a number of priority pollutants. The trained neural network is then used to predict the biodegradation kinetic constant for a new list of compounds, and the results have been compared with the experimental values and the predictionobtained from a linear group contribution method. It has been shown that the nonlinear group contribution method using neural networks is able to provide a superior fit to the training set data and produce a lower prediction error than the previous linear method.