A sample of 324 subjects were pre- and posttested and evaluated on the effect of burglary prevention literature that expressed a positive, negative, or mixed emotional message. The two types of conditions pertain to the message and those characteristic of receivers. In this study, the relationship between fear and communication turned out to be conditional rather than automatic: Side effects became apparent if and only if certain conditions were met. These conditions were formed by both message characteristics and receiver characteristics that proved to have differential consequences. The increase in general fear which followed exposure to negative communication was diminished by the general fear reduction resulting from exposure to positive communication. 6 tables and bibliography