A new method of hyperspectral imagery band selection is proposed based on artificial physics optimization algorithm. In this method, the combination between classes separability and information of band groups is used as fitness function, and subspace division is used to deal with hyperspectral imagery before band selection to reduce the correlation and redundancy of the output band of hyperspectral imagery. AVIRIS imagery is used for experiment with the proposed algorithm and other classical algorithms which are ant colony algorithm, genetic algorithm and particle swarm optimization, and it turns out that this algorithm is more effective in both band chosen performance and computing time consumption than others. 