Abstract: A microwave imaging approach based on real-coded genetic algorithm for a perfectly conduction cylinder in free space is presented. The contour of the cylinder is expressed by a shape function which is approximated by trigonometric series. The scattering problem is simulated by point-matching moment method; the inverse scattering problem is converted into a restrained optimization problem, that is, the coefficients of the trigonometric series are the variables to be optimized, and the optimization is made by minimizing the relative error between the measured amplitude of the scattered electric-field and the simulated one. The effectiveness of the image reconstruction method proposed in this paper is verified by an examples.