This study explores logical and empirical implications of friendship and enmity in world politics by linking indirect international relations (e.g., “the enemy of my enemy,”“the enemy of my friend”) to direct relations (“my friend,”“my enemy”). The realist paradigm suggests that states ally against common enemies and thus states sharing common enemies should not fight each other. Nor are states expected to ally with enemies of their allies or with allies of their enemies. Employing social network methodology to measure direct and indirect relations, we find that international interactions over the last 186 years exhibit significant relational imbalances: states that share the same enemies and allies are disproportionately likely to be both allies and enemies at the same time. Our explanation of the causes and consequences of relational imbalances for international conflict/cooperation combines ideas from the realist and the liberal paradigms. “Realist” factors such as the presence of strategic rivalry, opportunism and exploitative tendencies, capability parity, and contiguity increase the likelihood of relational imbalances. On the other hand, factors consistent with the liberal paradigm (e.g., joint democracy, economic interdependence, shared IGO membership) tend to reduce relational imbalances. Finally, we find that the likelihood of conflict increases with the presence of relational imbalances. We explore the theoretical and practical implications of these issues.