A longitudinal study was conducted to test whether the degree of campaign issue conflict portrayed in newspapers is linked to frequency of issue coverage in setting the agendas of prospective voters regarding important campaign issues. The research replicated and expanded agenda setting models of media effects, using 487 Indiana University students during the 1974 Bayh-Lugar senatorial campaign. Data on newspaper agendas were obtained from a content analysis of four newspapers during the two month period immediately prior to the election. Campaign issues were rank-ordered according to frequency of newspaper coverage and frequency of coverage combined with degree of conflict. Audience agendas were obtained on the basis of issues respondents were most concerned about, issues cited as reasons for supporting or opposing the candidates, and rankings of issues according to importance. Analysis of the results indicated that the audience's aggregate agenda rankings correlated significantly higher with the newspapers' aggregate agenda rankings based on issue freqency and conflict than with the newspapers' agenda rankings based on frequency alone, although the correlation was weaker than anticipated. (GW)