In this paper I develop a formal theory of campaign communications. Voters have beliefs about the quality of candidates in the different policy issues and about the issues' relative importance. Candidates spend time or money (TV ads, public speeches, etc.) in an effort to influence voters' decisions at the ballot. Influence has two simultaneous effects: (i) it increases the quality of the policy in the issue as perceived by the voters through policy/competence advertising and (ii) it makes the issue more salient through issue priming, thereby increasing the issue's perceived importance. A strategy is an allocation of influence activities to the different issues or topics. I show conditions under which candidates' strategies converge or diverge, which issues -if any -will dominate the campaign, and under what conditions candidates are forced to focus on issues in which they are perceived to be weak. The results are often conflicting with previous theories of campaigning but are able to explain a set of anomalies.