In this paper, multiuser detection in code division multiple access (CDMA) was performed by using neural network (NN) and parallel interference cancellation (PIC). Neural network is used as a front-end stage of one stage PIC circuit. PIC is a classical technique in multi user detection process and its bit error rate (BER) performance is not good in one stage for most of the applications. For improving its BER performance, generally multi stage PIC which has high computational complexity is used. In this study, we have got better BER performance than the three stage PIC receiver in AWGN channel, almost same BER performance with the three stages PIC receiver in the Rayleigh fading channel. © 2007 Springer Science+Business Media, LLC.