The purposes of this study were as follows: 1) to classify competitive swimmers on the basis of their stroke rate (SR) and stroke length (SL) by applying a factor analysis and a cluster analysis, 2) to derive discriminant functions by applying a multiple discriminant analysis of the stroke characteristics, and 3) to compare the race patterns of groups classified according to the results of the cluster analysis. 94 male swimmers who had participated in 100-m freestyle events held at the All Japan Swimming Championship from 2000 to 2004 were selected as subjects. The average value of their swimming velocity, SR, and SL in four swimming phases (15-25 m, 25-45 m, 57.5-75 m, and 75-95 m) were measured from videos captured. The subjects were classified according to their SR and SL by applying a factor analysis and a cluster analysis. Results showed that the competitive swimmers could be classified into Type1-4. Type1 showed high values of SR throughout the race, Type2 showed high values of SR in the first half and high values of SL in the second half, Type3 showed high values of SL in the first half and high values of SR in the second half, and Type4 showed high values of SL throughout the race. Although there was no difference in swimming records among the four groups, each group had different race patterns. These results suggest that the knowledge of race patterns might be useful to swimmers or coaches for planning race strategies.