Day-to-day variability and single-measurement reliability of selected isokinetic knee extension-flexion strength and endurance indices were assessed in 10 adult men and 8 adult women. On three occasions separated by at least 5 days, the subjects completed 4 reciprocal maximal voluntary contractions (MVC) at different angular velocities (1.05 rad.s-1 and 3.14 rad.s-1). The men also completed a muscular endurance test consisting of 30 reciprocal, MVC at 3.14 rad.s-1. Coefficient of variation, intra-class correlation coefficient and standard error of single-measurement scores support the continued use of gravity corrected peak torque (PT) and average peak torque (APT) as indices of isokinetic leg strength. Similarly, gravity corrected APT and total work should be the recommended indices of isokinetic leg muscular endurance in men. The results suggest that these isokinetic indices must be assessed using multiple day-to-day trial protocols adequately to describe performance capacity. Composite indices such as the ratio of Knee flexion to extension PT and fatigue measurements offer considerably reduced reliability and a greater potential for misinterpretation. The reliability of knee extension indices generally exceeds that of flexion indices. Similar variability and reproducibility of responses were observed between men and women and between reciprocal contractions performed at angular velocities of 1.05 rad.s-1 and 3.14 rad.s-1.