We study fairness and reciprocity in a Hawk-Dove game. This allows us to test various models in one framework. We observe a large extent of selfish and rational behavior. Our results are inconsistent with leading models in this field.