This study investigated the effect of introducing different interrepetition rest (IRR) periods on the ability to sustain maximum bench press throw velocity with a range of loads commonly used to develop upper-body power. Thirty-four physically active collegiate men (age: 21.5 ± 2.8 years; body mass: 75.2 ± 7.2 kg; height: 176.9 ± 4.9 cm) were tested during 2 consecutive weeks. During the first week, the maximum dynamic strength (repetition maximum [RM]) in bench press exercise was determined (RM = 76.7 ± 13.2 kg). The following week, 3 testing sessions were conducted with 48 hours apart in random order. In each day of evaluation, only 1 load (30%RM, 40%RM, or 50%RM) was assessed in the bench press throw exercise. With each load, subjects performed 3 single sets of 15 repetitions (15-minute interset rest) with 3 different sets configurations: continuous repetitions (CR), 6 seconds of IRR (IRR6), and 12 seconds of IRR (IRR12). The decrease of peak velocity (PV) was significantly lower for IRR12 compared with CR and IRR6 at least since the repetition 4. No differences between CR and IRR6 protocols were found until the repetition 7 at 30%RM and 40%RM and until the repetition 5 at 50%RM. The decrease of PV during the CR protocol was virtually linear for the 3 loads analyzed (r2 > 0.99); however, this linear relationship became weaker for IRR6 (r2 = 0.79–0.95) and IRR12 (r2 = 0.35–0.87). These results demonstrate that IRR periods allow increasing the number of repetitions before the onset of significant velocity losses.