Advancement in genetic algorithm (GA) optimization tools for design applications, coupled with techniques of soft computing, have led to new possibilities in the way computers interact with the optimization process. In this paper, the concept of goal-oriented GA has been used to design a tool for evaluating and optimizing various aspects of earth-to-air heat exchanger behavior. A new optimization method based on GA is applied as a generative and search procedure to optimize the design of earth-to-air heat exchanger. The GA is used to generate possible design solutions, which are evaluated in terms of passive heating and cooling of building, using a detailed thermal analysis of non air-condition building environment The results from the simulations are subsequently used to further guide the GA search to find the high-energy solutions for optimized design parameters. The specific problem addressed in this study is the sizing of earth-to-air heat exchanger in a non air-conditioned residential building. The developed algorithm is suitable for the calculation of the outlet air temperature and therefore of the heating and cooling potential of the earth-to-air heat exchanger system. This methodology is applicable to a wide range of design optimization problems like choice of building such as green house, solar house, or heating and cooling of buildings by mechanical system. (c) 2008 Elsevier Ltd. All rights reserved.