Effectiveness of stroke rehabilitation is important; there are about 62,000 strokes in Canada every year and more than 400,000 people living with long-term disability from stroke. Stroke survivors say that regaining walking ability is a top priority; but, reduced balance often limits community levels of mobility. Today’s stroke rehabilitation treatment decisions are guided by observed limitations to balance and walking, and not by brain measures. Differences in how much balance recovery someone achieves likely stems from the brain’s ability to recover after stroke. Functional near-infrared spectroscopy (fNIRS) is an imaging technology that can help us understand how the brain recovers balance after stroke, and may help identify those people who will benefit from an exercise intervention. fNIRS is special in that it can be used while people are moving during tasks like standing and balancing. At this point, it is unknown whether a recording with fNIRS can help us predict who will improve their balance, before starting an exercise intervention. So, the goals of our clinical trial, is to test which brain regions and how they interact, produce high levels of balance recovery. If we can find a fNIRS balance profile, instead of treating patients based on observed limitations, we will identify which individuals benefit from exercise programs. Given that stroke damages the brain, involving brain measures to guide treatment decisions could increase effectiveness of stroke rehabilitation.