A method based on multiple adaptive-network-based fuzzy inference system (MANFIS) is presented for the synthesis of electrically thin and thick rectangular microstrip antennas (MSAs). MANFIS is an extension of a single-output adaptive-network-based fuzzy inference system to produce multiple outputs. Six optimization algorithms, least-squares, nelder-mead, genetic, hybrid learning, differential evolution and particle swarm, are used to identify the parameters of MANFIS. The synthesis results of MANFIS are in very good agreement with the experimental results available in the literature. When the performances of MANFIS models are compared with each other, the best result is obtained from the MANFIS model optimized by the least-squares algorithm. 0 2008 Wiley Periodicals, Inc.