As the genetic algorithm is robust and independent of the specific optimization objectives, a method is presented that the best design of meridian line of progressive addition lenses is found through genetic algorithm. The polynomial coefficients of the meridian line are represented by binary character strings（chromosome） based on the smoothing equation. The multiple objectives of the power and astigmatism of the progressive addition lens are transformed into a single objective function through punish function and linear combination method. Selection, crossover and mutation operations are processed on the chromosomes to produce new generation population according to the objective evaluation function. The processing is repeated to find the best polynomial coefficients of the meridian line. Two optimizing examples are implemented and the optimized progressive addition lens is manufactured and tested. The results demonstrate that two optimized designs and the practical manufactured progressive addition lenses consist with each other.