Test–Retest Reliability of Maximal and Rapid Knee Extensor Force Production: Influence of Verbal Instruction and Trial Selection Criteria

Reliable assessments of maximal and rapid muscle strength are essential for evaluating neuromuscular performance and adaptations. This study examined the effects of verbal instruction and trial selection criteria on performance outcomes and intersession test-retest reliability during maximal isometric knee extensor testing. Twenty-three physically active, healthy adults completed three separate test sessions, performing isometric knee extensor contractions under three standardized verbal instructions: Hard and Fast (HF; "as hard and fast as possible"), Fast Only (FO; "as fast as possible"), and Hard Only (HO; "Gradual increase in force"). Rate of torque development (RTD) and impulse were calculated using three selection criteria: the trial with the highest peak torque (Max_Peak_Torque), the trial with the highest 200-ms impulse (Max_Impulse200), and the highest value for each time interval (Max_Composite_Interval). Both FO and HF instructions demonstrated good-to-excellent reliability, reflected by intraclass correlation coefficients (ICC) and within-subject coefficients of variation (CV_w-s) for RTD and impulse (ICC = 0.75-0.93, CV_w-s = 8%-15%), with improved reproducibility between sessions 2-3. Selection criteria strongly influenced test-retest reliability, with Max_Impulse200 and Max_Composite_Interval yielding higher ICCs and lower CV_w-s compared with Max_Peak_Torque. FO instruction consistently produced higher RTD and impulse values than HF (p ≤ 0.05). Peak torque increased across sessions, yet reliability remained excellent for both HF and HO (ICC > 0.90, CV_w-s < 7%). These findings indicate Fast Only instruction and selection of the Max_Impulse200 selection criterion may enhance reproducibility and magnitude of rapid knee extensor force assessments, especially after familiarization.