The purpose of this study was to determine the reproducibility of measures for maximum knee extensor and flexor concentric strength on an isokinetic dynamometer in individuals with mild and moderate osteoarthritis of the knee. Twenty eight female patients with 1(st) (n = 14) and 3(rd) (n = 14) grade unilateral knee OA volunteered for the study. Peak muscle torque of the knee extensors and flexors was measured using an isokinetic dynamometer at angular velocities of 90, 120 and 150 degrees/s. Test-retest reproducibility of maximal isokinetic torque was evaluated using intraclass correlation coefficient (ICC). Absolute reproducibility was assessed using Bland-Altman analysis, the standard error of measurement (SEM and SEM%) and the smallest real differences (SRD & SRD%). Inter-session relative reproducibility was found to be high in extensor and flexor muscles of individuals with mild OA, with the ICCs values ranging from 0.89 to 0.92 for the concentric knee extension and from 0.90 to 0.93 for the concentric knee flexion, while in the group of individuals with moderate knee OA the same values varied from 0.79 to 0.91 and from 0.75 to 0.93 for the concentric knee extension and flexion, respectively. The Bland-Altman analyses support the findings of high relative reproducibility for the two groups of patients in almost all measurements, except for the extension at 120 and 90 degrees/s and the flexion at 150 degrees/s conducted by the individuals with moderate OA. SEM and relative SEM values ranged from 6.7 to 9.4 Nm and from 14.4% to 21.5%Nm for the group with mild OA and from 4.4 to 7.7 Nm and from 14.6% to 21.8%Nm for the group with moderate OA, respectively. SRD and SRD% values across all movement speeds for both knee flexors and extensors ranged from 18.6 to 26.1 Nm and from 39.9% to 59.6%Nm for the mild OA group, respectively, while the same values for the moderate OA group ranged from 12.2 to 21.3 Nm and from 40.5% to 56.8%Nm, respectively. These results indicate that measurements of isokinetic performance at velocities of 90, 120 and 150 degrees/s provide acceptable reproducibility for evaluation of knee strength in individuals with mild OA of the knee. On the other hand, with regard to knee muscles isokinetic performance, testing of individuals with moderate OA should be conducted during periods of time when the symptoms of the disease subside.