Future design environments will necessitate improved management of the propagation and impacts of changes. To ascertain whether change prediction can assist in making better work prioritisation decisions, this paper develops a new simulation approach and applies it to a model of a complex aerospace product, which was elicited from industry. We use an accepted technique to generate potential change propagation trees and apply Monte Carlo methods to generate a sample space within which multiple scheduling policies could be evaluated and compared. The experiments reveal that poor coordination of change activity can result in significant process inefficiencies, that the potential for inefficiency increases for larger change networks, and that a modest ability to accurately predict change propagation in the specific case at hand could have a dramatic effect in reducing unnecessary rework. The experiments also suggest that the capability of predicting multiple steps of change propagation would provide only minimal additional improvement.