Classification of cauliflower resource was useful for heterosis and hybridization selection. 54 cauliflower inbred lines were analyzed with statistical methods of cluster analysis and the principal component analysis (PCA). The PCA results showed that, the six eigenvalues of cumulative variance proportion 70.024% were selected for evaluation of inbred line of cauliflower; Days to curd appearing, days to curd harvesting, leaf length, leaf width, plant height, lengthways diameter of curd, transverse diameter of curd, curd weight, curd shape, tight degree, leaf color, wax powder, leaf number inside, days to flowering and plant extent were morphological characteristics for parents screening. By cluster analysis under the results of PCA, 54 cauliflower inbreed lines were classified into 3 clusters. Characters of clusterⅠwere early maturity, little plant extent, narrow leaf with less wax, moderate weight and tight degree of curd; Cluster Ⅱ had the characters of middle maturity, moderate plant extent, sage green leaves with thick wax, and weighty curd with tight semicircular; Cluster Ⅲ was type of late maturity, high stalk, large plant extent, more length and width leaves with moderate wax, and oblate curd. Three clusters had obvious feature, so that it would be helpful for parent materials selection in breeding.