This investigation examined peak motor unit activity during sets that differed in resistance (50, 70, or 90% 1 repetition maximum [1RM]). Ten resistance-trained men (age, 23 ± 3 years; height, 187 ± 7 cm; body mass, 91.5 ± 6.9 kg; squat 1RM, 141 ± 28 kg) were assessed by electromyography (EMG) on the vastus lateralis and vastus medialis muscles in a randomized within-subject experiment consisting of 2 test visits: a drop-set day and a single-set day using only the 50% of 1RM intensity performed to failure. At the start of each day, subjects performed 2 submaximal repetition sets (50% 1RM × 10 repetitions and 70% 1RM × 7 repetitions). On the drop-set day, subjects performed 3 consecutive maximal repetition sets at 90%, 70%, and 50% 1RM to failure with no rest periods in between. On the single-set day, subjects performed a maximal repetition set at 50% 1RM to failure. Overall, the maximal repetition sets to failure at 50% and 70% 1RM resulted in higher peak EMG amplitude than during submaximal repetition sets with the same resistance. However, peak EMG amplitude was significantly (p ≤ 0.05) greater in the maximal 90% 1RM set than all other sets performed. When sets were performed to failure, ratings of perceived exertion (CR-10) did not differ over the intensity range of loads and suggests that perception is not capable of accurately detecting the actual amount of motor unit activation. The results of this investigation indicate that using higher external resistance is a more effective means of increasing motor unit activity than increasing the number of repetitions performed with lighter weights even when the end point is muscular failure. Accordingly, previous recommendations for the use of heavier loads during resistance training programs to stimulate the maximal development of strength and hypertrophy are further supported.