Purpose: Medial knee osteoarthritis (OA) is a mechanical condition affected by the amount of load on the knee joint. This load is determined by the magnitude of the external knee adduction moment (EKAM). Moreover, increased muscle co-contraction of the agonist and antagonist muscles around the knee joint have been reported to increase with knee OA with the possibility of a further increase in the loading at the joint. Exercises are a common treatment which have been shown to decrease pain, improves muscle strength and function. However, there is a paucity of literature on the effect of exercise regimens on both EKAM and muscle co-contraction and whether exercises increase or decrease loading on the joint. Previous studies strengthening either the quadriceps or hip adductors have found no significant change in EKAM which may be due to the type of exercises, the muscles exercises or method of progression. Therefore, the purpose of this study was to investigate the effectiveness of a neuromuscular training exercise programme of the hip and knee muscles on EKAM and the co-contraction of antagonist muscles in knee OA. Methods: This study was a pilot pre-post design study comparing the effects of a six-week exercise programme. A convenience sample of 19 participants were recruited from the waiting lists at a local Hospital. 3-dimensional kinematic (Qualisys OQUS, Gothenburg, Sweden) and kinetic (AMTI, USA) analysis were collected whilst walking. Surface electromyography (EMG) data were collected using a Noraxon Telemyo system (Noraxon, USA) and average normalised activation was recorded with electrodes placed on vastus lateralis (VL), vastus medialis (VM), biceps femoris (BF), and semitendonosis (ST) muscles on the affected side. Using the Biodex system 3 isokinetic dynamometer (Biodex Medical Systems, Shirley, N.Y., USA), the average peak torque of the knee flexors and extensors were assessed concentrically at 60°/s and isometrically at 45°, whereas the hip abductors were assessed isometrically at 0°. Clinically, the pain subscale of the Knee injury and Osteoarthritis Outcome Score (KOOS) questionnaire. The participants were assessed at baseline and after the six-week exercise programme and patient's attendance to the treatment sessions was recorded. Paired t-tests were used to determine the effects of the exercise programme on all outcome measures before and after the pilot exercise programme. Results: Fourteen participants (twelve women and two men, mean age 61.79±10.42 years) completed the exercise programme with good attendance rates (mean 5.36 (SD 0.84) sessions). The EKAM did not change significantly throughout the stance phase on the affected side though pain and muscle strength significantly improved (p<0.05). The mean difference (standard deviation) and p-values of EKAM during gait are as follows: in early-stance -0.01 (0.08), p= 0.57, in mid-stance -0.01 (0.05), p=0.70, and in late-stance -0.02 (0.06), p=0.21. Only muscle co-contraction between vastus-lateralis and biceps-femoris of the affected side significantly decreased in early and mid-stance (26.68 (34.84), p=0.01 and 12.68 (21.07), 0.04, respectively). Conclusions: The current pilot exercise programme was developed in an attempt to decrease the load on the knee joint. However, it can be seen from this preliminary data that EKAM did not increase or decrease significantly but muscle co-contraction on the lateral side of the knee joint decreased which may have reduced total loading on the joint. The reduced co-contraction may be one way in which pain was decreased in this exercise programme along with the increased muscle strength. This should be confirmed in a larger and longer RCT, as if the EKAM can be maintained with an exercise programme with increased adherence, rate of progression could be delayed.