What explains why the United Nations Security Council meets and deliberates on some armed conflicts but not others? We advance a theoretical argument centred on the role of conflict externalities, state interests and interest heterogeneity. We investigate data on the Security Council's deliberation on armed conflicts in the 1989–2019 period and make three key findings: (1) conflicts that generate substantive military or civilian deaths are more likely to attract the Security Council's attention; (2) permanent members are varyingly likely to involve the Security Council when their interests are at stake; and (3) in contrast to the conventional wisdom, conflicts over which members have divergent interests are more likely to enter the agenda than other conflicts. The findings have important implications for debates about the Security Council's attention, responsiveness to problems and role in world politics.