Background: Osteoarthritis of the knee involving predominantly the medial tibiofemoral compartment is common in older people, giving rise to pain and loss of function. Many people experience progressive worsening of the disease over time, particularly those with varus malalignment and increased medial knee joint load. Therefore, interventions that can reduce excessive medial knee loading may be beneficial in reducing the risk of structural progression. Traditional quadriceps strengthening can improve pain and function in people with knee osteoarthritis but does not appear to reduce medial knee load. A neuromuscular exercise program, emphasising optimal alignment of the trunk and lower limb joints relative to one another, as well as quality of movement performance, while dynamically and functionally strengthening the lower limb muscles, may be able to reduce medial knee load. Such a program may also be superior to traditional quadriceps strengthening with respect to improved pain and physical function because of the functional and dynamic nature. This randomised controlled trial will investigate the effect of a neuromuscular exercise program on medial knee joint loading, pain and function in individuals with medial knee joint osteoarthritis. We hypothesise that the neuromuscular program will reduce medial knee load as well as pain and functional limitations to a greater extent than a traditional quadriceps strengthening program. Methods/Design: 100 people with medial knee pain, radiographic medial compartment osteoarthritis and varus malalignment will be recruited and randomly allocated to one of two 12-week exercise programs: quadriceps strengthening or neuromuscular exercise. Each program will involve 14 supervised exercise sessions with a physiotherapist plus four unsupervised sessions per week at home. The primary outcomes are medial knee load during walking (the peak external knee adduction moment from 3D gait analysis), pain, and self-reported physical function measured at baseline and immediately following the program. Secondary outcomes include the external knee adduction moment angular impulse, electromyographic muscle activation patterns, knee and hip muscle strength, balance, functional ability, and quality-of-life. Discussion: The findings will help determine whether neuromuscular exercise is superior to traditional quadriceps strengthening regarding effects on knee load, pain and physical function in people with medial knee osteoarthritis and varus malalignment.