This article demonstrates the application of artificial neural network in multi-component analysis. Parameters were obtained after the BP network was trained with large amount of simulated data. Five organic toxins whose FTIR spectra are strongly overlapped were used to make the multi-component system. The relative standard deviation(RSD%), the percent standard error of prediction samples(SEP%) and the percent standard error of calibration samples(SEC%) were used for evaluating the ability of the neural network.