This study aimed to use the intermittent critical velocity (ICV) model to individualize intermittent exercise and analyze whether a fast-start strategy could increase the time spent at or above 95 %VO(2max) (t95VO(2max)) during intermittent exercise. After an incremental test, seven active male subjects performed three intermittent exercise tests until exhaustion at 100, 110, and 120 % of the maximal aerobic velocity to determine ICV. On three occasions, the subjects performed an intermittent exercise test until exhaustion at 105 % (IE105) and 125 % (IE125) of ICV, and at a speed that was initially set at 125 %ICV but which then decreased to 105 %ICV (IE125-105). The intermittent exercise consisted of repeated 30-s runs alternated with 15-s passive rest intervals. There was no difference between the predicted and actual Tlim for IE125 (300 ± 72 s and 284 ± 76 s) and IE105 (1,438 ± 423 s and 1,439 ± 518 s), but for IE125-105 the predicted Tlim underestimated the actual Tlim (888 ± 211 s and 1,051 ± 153 s, respectively). The t95VO(2max) during IE125-105 (289 ± 150 s) was significantly higher than IE125 (113 ± 40 s) and IE105 (106 ± 71 s), but no significant differences were found between IE125 and IE105. It can be concluded that predicting Tlim from the ICV model was affected by the fast-start protocol during intermittent exercise. Furthermore, fast-start protocol was able to increase the time spent at or above 95 %VO2max during intermittent exercise above ICV despite a longer total exercise time at IE105.