The usefulness of presidential debates to the electorate and to the total political system is evaluated in this paper. The paper first reports the results of opinion polls concerning the value of the 1976 debates and cites studies showing the types of information that people obtained from watching the debates. It then considers whether voters' learning gains affected the quality of their electoral decisions. Evidence is examined with regard to vote stability, party affiliation, image voting, other predispositional factors, and issue voting; it is concluded that, contrary to expectations based on the "limited effects" model (which stresses the tendency of mass political communication to strengthen already-made commitments), the debates made an important contribution to rational issue-based voting. The paper then summarizes the usefulness of debates for some, though not all, types of election campaigns and argues for maintaining the concept of presidential debates as a possibility for future elections. The final section of the paper examines the validity of claims about the functions that debates can serve in benefiting enduring governmental and political institutions and in enabling the administration that wins an election to govern effectively. (GW)