This study aimed to examine short-term resistance training effects of resting period length between sets on maximal number of repetitions and mean velocity over a moderate-intensity (60% of the maximum voluntary isometric contraction [MVIC]) set to failure on elbow-flexor muscles. The MVIC and surface electromyographic activity (sEMG) were also measured. Twenty-one untrained subjects were divided into 3 groups: short rest between sets (1 minute; SR), long rest between sets (4 minutes; LR), and nontraining control group (CG). The SR and LR performed 3 sets to failure in an arm-curl machine, 2 days per week for 5 weeks, with moderate loads (60-75% of the MVIC). The LR completed a significantly higher (31.6%, p < 0.05) total training volume than the SR. Both training groups enhanced the maximal number of repetitions to failure, with no significant differences in the magnitude of gains. The posttraining average velocity achieved by the SR at 40, 50, 60, 70, 80, and 90% of the total number of repetitions completed was significantly higher (p < 0.05) than the corresponding average velocity achieved on pretraining conditions, whereas no significant differences were observed in the LR. No significant changes in the MVIC or sEMG were observed in any group. We conclude that short-term elbow-flexor resistance training to failure, allowing 1 or 4 minutes of rest between sets, induces similar gains concerning local muscular endurance. Nevertheless, only the SR training approach reduced the rate of decline in the average repetition velocity during a set to failure. This can be of some importance in sport modalities in which not only the maximal number of repetitions (e.g., muscle endurance), but also a greater maintenance of high repetition velocities, may be critical for performance.