Abstract: Climate variability may introduce a range of air quality and health outcomes for a given emission control policy. While the influence of meteorology on air quality is well established, shifting climate baselines may affect the predicted response of air pollution to emission changes. When using air quality models for decision-making, it is important to quantify how various sources of uncertainty and variability affect the likelihood that a prospective emission control strategy will achieve a given air quality target. Here, we leverage an ensemble approach to assess how climate variability and change impact the sensitivity of ozone to nitrogen oxides (NO_x) emissions reductions. We consider three sources of variability: (1) inter-annual variability caused by natural forcings such as year-to-year variations in solar radiation, (2) unforced natural climate variability due to inherent stochasticity in the climate system, (3) external climate forcing uncertainty due to a range of possible future greenhouse gas emission trajectories. We use climate fields from a five-member initial condition ensemble of the Massachusetts Institute of Technology Integrated Global System Model linked to the Community Atmosphere Model (MIT IGSM-CAM) to drive offline simulations of the high-performance GEOS-Chem chemical transport model (GCHP). Analysis of the simulated ozone response to a 10% reduction in anthropogenic NO_x emissions reveals the largest spread in polluted urban areas. In these areas, the ensemble standard deviation and the inter-annual standard deviation of the ozone response can be on the order of 10-30% of the ensemble and multi-year means. In most areas, future climate variability does not induce a change in ozone photochemical regime. However, currently planned ozone control strategies in urban areas that experience a transitional or mixed photochemical regime are less robust to climate uncertainty.