BackgroundElevated lateral hamstring activity is often found in individuals with knee osteoarthritis during gait. These findings are based on maximal voluntary isometric contraction normalized EMG signals. This choice of amplitude normalization may contribute to differential activation of the hamstrings. Research questionThe objective was to determine lateral to medial hamstring root mean square activation ratios of individuals with medial compartment knee osteoarthritis compared to asymptomatic older adults during walking. The secondary objective was to determine whether this ratio differed between ipsilateral and contralateral knees in individuals with knee osteoarthritis. MethodsSurface electromyography of the hamstrings were acquired using standardized techniques from both limbs of 42 individuals with unilateral symptomatic medial compartment knee osteoarthritis and a random limb of 40 asymptomatic individuals during treadmill walking. Root mean squared amplitudes from the gait cycle were calculated. The lateral:medial activation ratio was computed. To address the first objective, an independent t-test was performed; both corrected and not corrected for walking velocity. A paired t-test was used for the second objective (alpha = 0.05). ResultsThe lateral:medial activation ratio was greater in the knee osteoarthritis group demonstrating a moderate effect size (p < 0.05, Cohens d = 0.73). The ipsilateral lateral:medial ratio was also greater than the contralateral (p < 0.05) in the knee osteoarthritis group, showing a low to moderate effect size (Cohens d = 0.53). SignificanceThe activation ratio of the lateral and medial hamstrings during treadmill walking was unique to the symptomatic leg of individuals with medial compartment knee osteoarthritis. The ratio showed a bias toward greater lateral hamstring activation in the symptomatic leg. While considerations such as the impact of subcutaneous tissue differences between electrode sites should be considered when interpreting un-normalized electromyograms, this technique may be useful in integrating electromyography into clinical knee osteoarthritis functional assessments without the requirement of maximal voluntary isometric contraction-based amplitude normalization.