The Validity and Reliability of the Individual Load–Velocity Profiles in the Deadlift and Hip Thrust Exercises in Men and Women

Muscular strength assessment is fundamental for training optimization. While one-repetition maximum (1RM) testing is valid, its practical use is limited, making velocity-based training (VBT) a viable alternative for estimating maximal strength and monitoring performance using submaximal loads. This observational, cross-sectional validity and reliability study aimed to: (i) assess the absolute and relative reliability of movement velocity at submaximal loads in the deadlift and hip thrust, (ii) evaluate the validity and reliability of 1RM estimations using LVP in these exercises, and (iii) compare reliability metrics between sexes. Thirty-two physically active adults voluntarily participated in the study, including sixteen men (age: 25.63 ± 3.79 years; body mass: 75.79 ± 8.64 kg; height: 175.81 ± 7.34 cm) and sixteen women (age: 25.06 ± 5.37 years; body mass: 61.94 ± 4.25 kg; height: 165.06 ± 5.72 cm). Strength-trained participants performed familiarization and two experimental sessions involving progressive load tests at various percentages of 1RM. Movement velocity was monitored using a validated linear velocity transducer. Statistical analyses included two-way ANOVA, Bland–Altman plots and intraclass correlation coefficients (ICCs). The results showed a systematic overestimation of 1RM derived from load–velocity profiles, which was moderate for the deadlift (mean difference ≈ +2.7 kg; p < 0.001) and small-to-moderate for the hip thrust (≈+3.3 kg; p = 0.002). Strong associations were observed between actual and estimated 1RM in both exercises (R2 > 0.92). Bland–Altman analyses revealed smaller bias but narrower limits of agreement for the deadlift compared with the hip thrust, indicating greater variability in hip thrust estimations. Test–retest analyses demonstrated good relative reliability of movement velocity in men across all loads (ICC > 0.75; CV < 8.2%), whereas women exhibited lower reliability, particularly at moderate-to-high loads in both exercises. Despite limited validity for precise 1RM estimation, load–velocity profiles showed good reproducibility, supporting their use for longitudinal performance monitoring rather than exact strength prediction.