College undergraduate smokers and nonsmokers were exposed to a persuasive anti-smoking communication. Measures were taken of their emotional arousal, attitudes toward smoking, perceived attitudes of family and friends, and intentions to engage in a variety of smoking-related activities.
A follow-up measure of their behavior was taken one to two weeks af tar-ward. It was predicted that the influence of emotional state upon intentions and behavior would be indirect, being mediated through attitudes toward smoking. Attitudes and normative (family and friends') beliefs were
expected to predict intentions, which were expected to predict actual behaviors. The results supported Leventhal (1970) and Fishbein and Ajzen (1975). Emotional state was independent of attitudes, perceived attitudes of family and friends, and all but one intention. Attitude-intention consistency
was observed for all responses, while intention-behavior consistency was limited to two responses. Family attitudes predicted intentions to cut down on the amount smoked and stop smoking altogether. Intention-behavior inconsistency was discussed in terms of barriers that may inter fere with
accomplishing one's intentions, and measurement issues which may distort the observed relationship between verbal intentions and subsequent actions.