The brake friction materials in an automotive brake system are considered as one of the key components for overall braking performance of a vehicle. Temperature sensitivity of friction materials has always been a critical aspect while ensuring their smooth and reliable functioning, and that sensitivity need to be constantly optimized. The performance of friction materials at elevated temperatures is defined by their fading performance.In this paper,a group of non-asbestos organic based friction materials containing different relative amounts of the ingredients of aluminosilicate fiber, steel fiber, phenolic resin and Cu powder mostly effecting on the tribological properties of designed brake material were manufactured by compression molding. Dry sliding friction characteristics of composites were tested on a model JF150D-II pad-on-disk type friction tester. A model of feed-forward artificial neural networks (ANN) consisting of four input neurons, six output neurons and one hidden layer, was used for the analysis and prediction of the correlation between material components and friction performance. The input parameters of ANN were the contents of four main components as aluminosilicate fiber, steel fiber, phenolic resin and Cu powder. The outputs were the friction coefficients of brake material against cast iron at six different operating tempetures from 100°C to 350°C. Based on ANN model trained successfully by 25 pieces samples, genetic algorithm(GA) was used to optimize the input parameters of compositions of brake material with the goal of minimizing fluctuation of friction coefficients at 100°C to 350°C. The optimum composition of brake material was obtained. The friction experiments of optimized material showed excellent stability of friction coefficients for automotive brake material as test temperatures increased from 100°C to 350°C. The neural network has impressive potential for solving time-consuming problems for the design of automobile brake material.