This study examined whether there is a transfer effect to other joint angles and the angular specificity of muscle fatigue after 6 weeks of isometric training of the vastus medialis. Twenty subjects were randomly assigned to 30° and 90° knee flexion groups and were trained at 80% maximal voluntary isometric contraction(MVIC) three times a week for 6 weeks. The pre-and post-training values of the 80% holding time(endurance time) of MVIC, the Fatigue Index(FI), and the MVIC at 30°, 60°, and 90° were compared. After isometric training for 6 weeks, in the 30° knee flexion group, FI decreased significantly(p<0.05) only at 30°, which was the training angle; there was no change at other angles. By contrast, in the 90° knee flexion group, FI decreased significantly(p<0.05) at both the trained angle and at the other angles, indicating a transfer effect of training. MVIC did not increase significantly(p<0.05) at any trained angle in either the 30° or 90° knee flexion groups after 6 weeks of isometric training, neither did the 80% holding time of MVIC differ significantly compared with pre-training in either group. These findings suggest that training at 90° of knee flexion is more effective than training at 30° of knee flexion for obtaining a training transfer effect on muscle fatigue in the vastus medialis.