This paper analyzes the methods of city security evaluation, constructs city community development quality evaluation index system By applying the theory of neural network into the comprehensive evaluation of city development level ,and the comprehensive analysis of city community quality, evaluates on time scales city community quality and its trend, provides theoretical basis for city security evaluation and management. And finally gives the application examples. Introduction City is a region of information center and transportation hub with politic, economy, culture, education science and technology, its development can often determine the regional economic development pattern and the spatial distribution of the allocation of resources. How to use varieties of effective ways to the comprehensive evaluation and master urban development trend, becomes very important point. Professor Wang Zongjun, puts forward the fuzzy evaluation method with a set of multi-objective and multi-level theory based on the theory of fuzzy of city comprehensive development level. But this method is lack of self-learning ability, and can't get away from the decision-making process of randomness and evaluation experts subjective uncertainty and fuzziness of knowledge. Neural network theory is applied to evaluate the city development level, provides theoretical basis for city security evaluation and management. Neural network characteristic Neural network is a dynamic system with directed the graph topology structure, it reacts and completes the information processing through a variety of input information, which has dynamic processing and self-learning, self-organizing, adaptive and nonlinear characteristics .Radial Basis Function (RBF) neural network is a feed-forward network with good performance, and it can decide the appropriate network topology based on different issues, with a high approximation precision, a small-scale of network training, fast learning speed and non-existence of local minima problems . The structure of RBF neural network consists of three layers: the input layer, hidden layer and output layer. The structure of topology is showed in Fig1.