The primary aim was to evaluate the effect of a dosed walking program on knee pain for patients with severe knee osteoarthritis (OA). Secondary aims evaluated the effects on cardiovascular health, function and quality of life.Participants with severe knee OA and increased cardiovascular risk were randomly assigned to a 12-week walking program of 70 min/week of at least moderate intensity, or to usual care. The primary outcome was knee pain (0-10). Secondary outcomes were of cardiovascular risk including physical activity, blood pressure, blood lipid and glucose levels, body mass index and waist circumference; WOMAC Index scores; physical function; and quality of life.Forty-six participants (23 each group) were recruited. Sixteen participants (70%) adhered to the walking program. Intention to treat analysis showed no between-group difference in knee pain. The walking group had increased odds of achieving a healthy systolic blood pressure (OR = 5.7, 95% CI 1.2-26.9), and a faster walking speed (Mean Difference (MD) = 0.12 m/s, 95% CI 0.02-0.23). Per protocol analysis based on participant adherence showed the walking group had more daily steps (MD = 1345 steps, 95% CI 365-2325); more time walking (MD = 18 min/day, 95% CI 5-31); reduced waist circumference (MD = -5.3 cm, 95% CI -10.5 to -0.03); and increased knee stiffness (MD = 0.9 units, 95% CI 0.07-1.8).Patients with severe knee OA prescribed a 12-week walking program of 70 min/week may have had cardiovascular benefits without decreasing knee pain. Australian New Zealand Clinical Trials Registry ACTRN12615000015549.