The study's hypothesis predicts that response generalization is expected in persons with an external risk perception, those with a high risk assessment, persons who feel police protection is inadequate, and individuals who have taken preventive measures in the past. Confrontation with a crime prevention message could easily form an incitement to go further, resulting in increased agreement with more extreme preventive reactions. To test the hypothesis, 324 residents of one-family homes in a city near Amsterdam, Holland, were randomly sampled. Subjects were given information on burglary prevention intended to provoke innocent preventive reactions such as locking doors and windows when leaving the house and installing a burglar-proof hinge. Those exposed to the crime prevention message showed high communicative and emotional involvement with crime, or a high-risk assessment. The paper discusses implications of response generalization for crime prevention campaigns and proposes strategies for avoiding these side-effects. Tables and 31 references. (Author abstract modified)