Abstract: Vehicular Ad hoc Network (VANET) is a promising technology for future smart vehicles systems and an essential component of Intelligent Emergency System (IES). The IES includes a wide range of modern technologies such as Global Positioning System (GPS), digital maps, video cameras, sensing devices and the wireless communication devices. It provides necessary information about the condition of the roads in time for drivers and traffic management systems to improve traffic efficiency, reduce traffic congestion, waiting times and fuel consumptions. Design and implement an IES which automatically controls the encryption of the Cooperative Awareness Messages (CAMs) according to the priority associated with CAMs exchanged between emergency vehicles and Road Side Units (RSUs). The CAMs sent from the emergency vehicles to RSUs be signing using a Secure Hash Algorithm (SHA-2) to distinguish them from normal messages issued from other vehicles. The IES uses the features extracted from the trace file that describes the normal and urgent behavior in the VANETs. The type and the number of features have an important role in increasing the classification accuracy rate and decreasing false alarms, especially False Negative Rate (FNR). In this study, the process of classification of urgent records by using (self-organizing map, feed-forward neural network and Elman neural network). The proposed system is based on a program written by MATLAB R2015a. Our selection used for design and programming the proposed system. These algorithms have already been employed to solve the problem because of its importance in saving time and effort as well as providing high results accuracy in quick time unlike other programming languages. The result is clear in overall system in each technique in SOM accuracy degree 99.5% and FNR 0% while FFNN accuracy degree 99.3% and FNR 0.84211% for number of features 16.