A genetic algorithm (GA) approach is proposed for the reconstruction of static images in electrical impedance tomography (EIT). Genetic algorithms can be demonstrated to possess several advantages over more conventional "gradient-based" techniques. In particular, they are implicitly parallel and realize a good compromise between "exploration" and "exploitation", thus being more robust against the problem of false minima. The results of GA-EIT in numerical experiments are presented, compared to those obtained by other, more established inversion methods, such as the modified Newton-Raphson method and the double-constraint method. The GA approach is relatively expensive in terms of computation time and resources, requiring (for example) from several minutes to tens of minutes on a Pentium Pro 200-based machine for normal-size EIT problems. This currently limits the applicability of GA-EIT to the field of static imaging. However, in light of the development trend in the field of computing, an extension to real-time dynamic imaging applications is not inconceivable in the near future.