Strategic air traffic flow management requires predictions of airport departure and arrival capacities several hours into the future. It is difficult to do this well because of weather forecast uncertainty. However, if the measurement of prediction uncertainty is provided, strategies can be developed to account for it. In this paper, we examine the potential for using ensemble weather products to quantify uncertainty in airport capacity prediction. An improved prediction model from previous studies that utilizes weather forecast variables to predict runway configuration and capacity is validated and then applied to each ensemble member, so that the distribution of possible capacity outcomes and prediction confidence can be obtained. Our results suggest that the performance of the prediction model has been improved, and that the forecast variation among ensemble members effectively captures the possible variations in future arrival capacity for certain airports.