Abstract: Clustering analysis is one of primary techniques in the field of data mining.It is an unsupervised mode of pattern recognition.Clustering analysis is a division of data into similarity groups according to given rules.K-means algorithm is a classical clustering algorithm.In the paper K-means’ advantage and weakness is analyzed.A novel clustering algorithm based on artificial fish swarm is proposed then a new mixed clustering algorithm is obtained from the combination of K-means and clustering algorithm which is based on artificial fish swarm.The result of the experiment shows it is a kind of efficient algorithm.