Abstract: As n-variable Reed-Muller(RM) logic functions have 3ⁿ mixed polarities, corresponding with 3ⁿdifferent expansions, a new method is proposed to search for the best polarity with respect to area by genetic algorithm-discrete ternary particle swarm optimization(GA-DTPSO) algorithm. A mathematic model for area is established for mixed polarity XNOR/OR circuits according to the form of circuit. Polarity conversions method is improved and applied in the polarity conversion of mixed polarity XNOR/OR circuits. GA-DTPSO algorithm by introducing discrete ternary particle swarm optimization to genetic algorithm is presented to improve the accuracy. Combining polarity conversions and the area mathematic model, the best polarity of mixed polarity XNOR/OR circuit is found by GA-DTPSO algorithm. Experimental results on fourteen MCNC and ISCAS benchmarks illustrated the effectiveness of the proposed method for area optimization problems of mixed polarity XNOR/OR circuits.