Isokinetic strength tests are probably the most common measurement tools for estimating muscle function in sport setting. However, there is little scientific evidence justifying their use as accurate tools. For that reason, the main purpose of this study was to examine the absolute reliability of isokinetic functional and conventional strength ratios as well as concentric and eccentric isokinetic knee flexion and extension peak torque (PT) of the knee in recreational active young adults.

Fifty two participants performed 3 measurement sessions to record the functional and conventional strength ratios and concentric and eccentric knee flexion and extension PTs with a 72-96hours rest interval between consecutive sessions. The inter-session absolute reliability was measured through change in the mean between measures (expressed as percentage value [CM]), typical percentage error (% within-subjects variation [CVTE]), and intraclass correlations (ICC).

Poor absolute reliability scores were found for conventional (CM=0.2%-0.4%; CVTE=16%-18%; ICC=0.6-0.7) and functional (CM=−2.1%-0.2%; CVTE=18%-20%; ICC=0.3-0.7) knee strength ratios. Likewise, moderate absolute reliability scores were observed for PT values independently of the knee movement (flexion and extension), type of muscle contraction (concentric and eccentric) and angular velocity (60°/s and 180°/s).

An observed change greater than ≈ 12% and ≈ 17% for PT scores (independently of knee movement and muscle contraction) and isokinetic ratios (conventional and functional), respectively, from baseline scores would suggest that a real change in strength balance was likely and not because a measurement error. These absolute reliability scores could be sensitive enough to detect large changes, as showed after completing rehabilitation programs for subjects injured.