R recent years have brought an amplification of voices favoring Presidential primaries, advocates of which base their argument on empirically untested assumptions that those who vote in primaries are sufficiently large in number and diverse in political interest to represent preferences of Republicans and Democrats across nation. Indeed, James W. Davis argues that primaries already have become form of 'plebiscitarian democracy' . . . [and that for out-party] national party conventions are becoming ratifying bodies that approve national as expressed in presidential primaries.' Davis's thesis that primaries determine out-party's nominee, i.e., that they are instruments of direct democracy, has been challenged effectively elsewhere But his contention that presidential primaries reflect the national choice has not received systematic attention. The purpose here is to measure extent to which primaries reflect by conjoining two criteria of with primary election data relevant to them. The component in popular choice clearly connotes participation by large numbers of people. The amount of available to a primary electorate depends upon number of serious presidential candidates in primary, presence or absence of important issues in primary campaign, and extent to which candidates join, or oppose, each other on issues.3 It is