Skiing technique is required to cope with topographical changes in snow ground condition. Besides, muscle strength is required to endure muscle fatigue in a high-speed environment, especially during a critical phase such as consecutive ski turns. Thus, objective evaluation is important for improving skiing technique and obtaining high performance. In this study, we analyzed the performance of actual skiing with different technique levels, focusing on leg muscular activities during ski control and postural control, in relation to the movement skills. We prepared the feature parameters estimated from surface EMG signals and knee joint angles at each turn, then, analyzed the performance by a cross-correlation for evaluating the relationships between the feature parameters. The results showed that the standard deviations of parameters in high-performance skiers were small. Cluster analysis with the k-means method, further, revealed muscle activity patterns in a three-dimensional scatter distribution. This approach leads to an objective evaluation on a ski ground that allows effective training and protects sport injury.