The invention discloses a hyperspectral remote-sensing image waveband selection method based on a genetic bee colony algorithm, is high in classification accuracy and obviously improves the classification accuracy of a hyperspectral remote-sensing image waveband selection method. Compared with a genetic algorithm, an ant colony algorithm and an artificial bee colony algorithm, the method disclosed by the invention has an obvious advantage on an aspect of classification performance. The genetic algorithm has high global search ability, but the rate of convergence of the genetic algorithm becomes slow along with the increase of iteration times, and the artificial bee colony algorithm has defects on an aspect of global search but is high in algorithm rate of convergence. By use of the method, the genetic algorithm is combined with the bee colony algorithm to form complementary advantages, and the algorithm simultaneously has the advantages of high global search ability and high rate of convergence. Therefore, on a premise that high classification accuracy is guaranteed, algorithm execution efficiency is high.