Cerebrovascular reactivity (CVR) mapping with MRI is emerging as a useful metric of vascular health. Tracking natural fluctuations in CO2 (a vasodilator) during a resting state fMRI scan has been proposed as a method to map CVR, but it is uncertain whether this method is reliable. We propose adding a two-minute breathing challenge prior to a resting state acquisition to transiently reduce CO2 levels. Here, we tested the feasibility of this method in a small pediatric cohort. The breathing challenge produced more robust CVR maps in both controls and children with cerebral palsy compared to resting state data alone.