Background: The aim of this study was to verify the reproducibility of a resistance training protocol in the bench press (BP) exercise, based on traditional recommendations, analysing the effect of the muscle fatigue of each set and of the whole exercise protocol. Methods: In this cross-sectional study, thirty male physical education students were divided into three groups according to their relative strength ratio (RSR), and they performed a 1RM BP test (T1). In the second session (T2), which was one week after T1, the participants performed a BP exercise protocol of three sets with the maximum number of repetitions (MNR) possible to muscle failure, using a relative load corresponding to 70% 1RM determined through the mean propulsive velocity (MPV) obtained from the individual load–velocity relationship, with 2 min rests between sets. Two weeks later, a third session (T3) identical to the second session (T2) was performed. The MPV of each repetition of each set and the blood lactate level after each set were calculated, and mechanical fatigue was quantified through the velocity loss percentage of the set (% loss MPV) and in a pre-post exercise test with an individual load that could be lifted at ~1 m·s−1 of MPV. Results: The number of repetitions performed in each set was significantly different (MNR for the total group of participants: set 1 = 12.50 ± 2.19 repetitions, set 2 = 6.06 ± 1.98 repetitions and set 3 = 4.20 ± 1.99 repetitions), showing high variation coefficients in each of the sets and between groups according to RSR. There were significant differences also in MPVrep Best (set 1 = 0.62 ± 0.10 m·s−1, set 2 = 0.42 ± 0.07 m·s−1, set 3 = 0.36 ± 0.10 m·s−1), which significantly reduced the % loss MPV of all sets (set 1 = 77.4%, set 2 = 64%, set 3 = 54.2%). The lactate levels increased significantly (p < 0.05) (set 1 = 4.9 mmo·L−1, set 2 = 6 mmo·L−1, set 3 = 6.5 mmo·L−1), and MPV loss at 1 m·s−1 after performing the three sets was 36% in T2 and 34% in T3, with acceptable intrasubject variability (MPV at 1 m·s−1 pre-exercise: SEM ≤ 0.09 m·s−1, CV = 9.8%; MPV at 1 m·s−1 post-exercise: SEM ≤ 0.07 m·s−1, CV = 11.7%). Conclusions: These exercise propositions are difficult to reproduce and apply. Moreover, the number of repetitions performed in each set was significantly different, which makes it difficult to define and control the intensity of the exercise. Lastly, the fatigue generated in each set could have an individual response depending on the capacity of each subject to recover from the preceding maximum effort.