Abstract：BP is the most commonly used artificial neural networks，but it sufers from extensive computations，rel— atively slow convergence speed and other possible weaknesses for complex problems．Genetic Algorithm (GA)has been successfully used to train neural networks，but often with the result of exponential computational complexities and hard implementation．Hence Ant Colony Optimization(ACO)is used to train BP in the paper．The eficiency of BP trained with ACO is compared with those of BP and．_BP trained with GA based on the identification of the same chaotic system．Comparison based on the searching precision and convergence speed of each method show that BP trained with ACO is dominant and effective to identify chaotic system．