Investigations (DiRenzo, 1968, 1971) of the 1964 and 1.968 presidential elections supported the hypothesis thac voting behavior is in part a function of some measure of psychological identification between the political elector and the political candidate. These data also indicated a differential and significant pattern for the supporters of the two major political parties on the basis of dogmatism and personality structure. Subsequent analyses in the longitudinal study of these questions were conducted in each of the presidential elections since 1972. Method.-Nonprobability samples of voters were administered either Form E (complete version) or Form D (abbreviated version) of the Rokeach Dogmarism Scale as part of an interview schedule concerned with interest and activity in the respective presidential elections. Details on procedures can be found in DiRenzo ( 1968, 197 1 ) . Results.-Results from each of the subsequent elections ( 1972, 1976, 1980, 1984) are in the hypothesized directions; however, in only one election ( 1976) were correlations between personality and party/candidate preferences statistically significant. Our tentative interpretation for the inconsistent results is thac the respective candidates and issues involved did not offer sufficiently specific and differentiated stimuli to permit a significant interaction of the variables. It appears, moreover, that the only typical election, one without any unusual political event, occurred in 1976. Each of the other elections offered a confounding political situation (Watergate in 1972, a third-party candidacy in 1980, and the landslide in 1984). One candidare (Caner) played a cenrral role in each of these three elections. Further analysis concerns the clarification of the interaction of the ideological and substantive contents of each elecrion, while the study of subsequent elections will focus on a greater specification of the political stimuli involved.