To what extent does the separation of powers affect congressional roll call voting behavior? To answer this question, I offer a strategic model of congressional decision making that asserts members of Congress pursue public policy goals when casting roll call votes. From the equilibrium predictions of a formal model, I generate testable hypotheses by computing the expected net amount of sophisticated (nonsincere) congressional behavior given changes in decision context. I test the predictions of the theoretical model with data from all civil rights roll call votes from the 83d to the 102d Congress. The results demonstrate that both the other legislative chamber and the Supreme Court profoundly constrain House members and senators when casting roll call votes. This is strong evidence of the importance of policy outcomes to members of Congress when voting on the floor.