Effect of different load-velocity profile testing procedures on 1-RM load and velocity at 1-RM in the Smith machine bench press

This study examined whether different load-velocity profile (LVP) testing protocols affect one-repetition maximum (1-RM) load in the Smith-machine bench press and, secondarily, velocity at 1-RM (V 1−RM ), expressed as mean velocity (MV), mean propulsive velocity (MPV), and peak velocity (PV). 15 men performed three LVP protocols in separate sessions, differing in load increment patterns. Load increments were based on execution velocity (10 kg for MPV>0.5 m·s −1 or 2.5–5 kg for MPV<0.5 m·s −1 LVP VB ), fixed percentages of 1-RM (20, 40, 60, 80, 90% of 1-RM LVP PB ), or two loads (50 and 80% of 1-RM LVP 2−POINT ). Outcomes were analyzed using repeated-measures ANOVA (α = 0.05), intraclass correlation coefficient (ICC), and Bland-Altman analyses. For 1-RM, a large but non-significant protocol effect was observed (p = 0.051), with the highest value for LVP 2−POINT (93.3 ± 16.8 kg). Reliability was excellent (ICC = 0.972). Bland-Altman analyses showed small biases but moderate-to-wide limits of agreement (LoA): LVP PB vs LVP VB , 1.37 kg (−4.71 to 7.45 kg), LVP PB vs LVP 2−POINT , −1.17 kg (−8.31 to 5.97 kg), and LVP 2−POINT vs LVP VB , 2.53 kg (−6.39 to 11.46 kg). For V 1−RM , protocol significantly affected PV (p = 0.020), with higher values for LVP 2−POINT (0.39 ± 0.09 m·s −1 ) than LVP VB (0.33 ± 0.09 m·s −1 ), while no differences were detected for MPV or MV. Reliability was poor-to-moderate (ICC = 0.050–0.641), with wide LoAs for all velocities. Overall, 1-RM can be reliably assessed across LVP protocols to discriminate inter-individual differences; however, moderate-to-wide LoAs indicate substantial intra-individual variability, limiting their interchangeability at the individual level. The LVP 2−POINT may be a reliable alternative to traditional 1-RM assessment, whereas V 1−RM appears less reliable and highly protocol-dependent.