We apply panel estimation methods to test how the incidence of war in neighbourhood countries affects the incidence of war in the domestic country. We find positive and significant evidence of "spatial conflict spillover": higher incidence in the neighbourhood increases domestic country involvement in war. We also provide some new evidence on other correlates of war. Particularly interesting are the results concerning the effect of democracy. An increase in the level of democracy increases the incidence of interstate war and reduces the incidence of civil war. However, this latter effect kicks in only past a certain initial level of democracy.