Do audience costs have to be extremely large in order to credibly signal resolve and affect international crises? Existing theoretical work on audience costs suggests an affirmative answer, and recent empirical work on audience costs focuses on whether a leader can generate such large audience costs as to create a commitment to fight where no such commitment previously existed. We analyze a richer crisis bargaining model with audience costs and find that (1) audience costs can have war-reducing effects on incomplete-information crisis bargaining through a noninformative, bargaining-leverage mechanism and (2) audience costs can have war-reducing effects even when such large audience costs are not being generated as to create a commitment to fight where no such commitment previously existed. Even more limited audience costs can have war-reducing effects in international crises. We discuss how the bargaining-leverage mechanism is consistent with a number of prominent historical cases. © The Author(s) 2012.