An asymmetry exists in the empirical literature on sophisticated voting in Congress. All studies that find supporting evidence of sophisticated voting have been only piecemeal—that is, they examine only one or a few roll calls. In contrast, the studies that systematically study many roll calls conclude that sophisticated voting is, at best, very rare. We are aware of three such systematic studies—those by Poole and Rosenthal (1997), Wilkerson (1999), and Ladha (1994). While Ladha's study has gone relatively unnoticed, we reexamine his results and explain why they may be the most important of all empirical work on sophisticated voting. In addition, we introduce a theoretical model, and we show how it, along with some subtle aspects of the rules for voting in the House and Senate, provides a rational-choice explanation for the lack of sophisticated voting in Congress.