In a Wireless Sensor Network when an event is detected, the network traffic increases. It in turn increases the flow of data packets and congestion. Congestion in Wireless Sensor Network plays a vital role in degrading the performance of the network. Hence it necessitates, developing a novel technique to control congestion. In this paper, soft computing based congestion control technique is proposed. Fuzzy logic and neural network are the soft computing tools used for estimating the packet drop. The performance of the proposed technique is evaluated using Accuracy. From the results, it is proved that neural network based congestion control technique provides better results than fuzzy based congestion control technique.