Making others especially fearful of what can happen if they ignore the advice given in a health-threat warning (fear appeal) encourages their acceptance of the message most of the time, though occasionally this strategy backfires. Current formulations of fear appeals are unable to reconcile these findings satisfactorily. The drive model notion that high fear arousal leads to "defensive avoidance" is not well supported by previous research; the parallel response model assertion that fear control behavior interferes with adaptive responding to health-threat warnings remains untested; and protection motivation theory is more concerned with isolating important components of health-threat warnings than with accommodating these apparently discrepant findings. A new conceptualization of health-threat warnings emphasizing the importance of perceived control over health is presented. Its essence is that health-threat warnings can diminish or enhance perceived control over health by lowering or raising perceptions of response and personal efficacy. It is shown that a model incorporating these two control constructs is capable of generating strategies for enhancing the persuasive impact of health-threat warnings in situations where other models encounter difficulties. (Author)