THE RESEARCHER PRESENTED THREE SITUATIONAL ENCOUNTERS BETWEEN A POLICE OFFICER AND A STUDENT TO 178 LAW ENFORCEMENT (LE) MAJORS AND 153 NON-LE MAJORS, ALL MALES, AT EASTERN KENTUCKY UNIVERSITY. IN THE ACCOUNTS, EITHER THE POLICE OFFICER OR THE STUDENT RECEIVED AN ACCIDENTAL INJURY THAT WAS EITHER MILD OR SEVERE. LE AND NON-LE MAJORS DID NOT DIFFER IN THEIR EVALUATIONS OF STUDENTS, ALTHOUGH LE MAJORS MADE MORE FAVORABLE EVALUATIONS OF THE TERM 'POLICEMAN' THAN DID NON-LE MAJORS. THE STUDY FOUND SUPPORT FOR THE PREDICTION THAT DEFENSIVE ATTRIBUTIONS OCCUR AS FUNCTIONS OF RELEVANCE. RESULTS CONTRARY TO PREDICTIONS SHOWED THAT RELEVANCE OF THE VICTIM DID NOT DIFFERENTIALLY INFLUENCE ATTRIBUTIONS TO CHANCE, INCREASED SEVERITY OF THE INJURY DID NOT MAGNIFY THE DEFENSIVE ATTRIBUTION PATTERN THAT OCCURRED AS A FUNCTION OF RELEVANCE, AND INCREASED SEVERITY DID NOT ACCENTUATE ATTRIBUTIONS TO CHANCE AS A FUNCTION OF THE RELEVANCE OF THE VICTIM. A PERSON WAS NOT EVALUATED LESS FAVORABLY OR SEEN AS MORE RESPONSIBLE FOR AN ACCIDENT WHEN HE WAS THE VICTIM OF AN ACCIDENT THAN WHEN HE WAS NOT INJURED. THE JUST WORLD MAKES NO PREDICTIONS OF ATTRIBUTIONS TO CHANCE AS A FUNCTION OF RELEVANCE. INCREASED SEVERITY OF THE INJURY DID NOT RESULT IN MORE NEGATIVE EVALUATIONS OF THE VICTIM OR INCREASED ATTRIBUTIONS OF RESPONSIBILITY TO HIM. FINALLY, INCREASED SEVERITY WAS NOT ASSOCIATED WITH DECREASED ATTRIBUTIONS TO CHANCE. RESULTS INDICATE THAT THE LIKELIHOOD OF ENCOUNTERING BIASES EXISTS, ALTHOUGH NO INCREASED BIAS WAS ASSOCIATED WITH INCREASED SEVERITY. THE STUDY RECOMMENDS FURTHER REFINEMENT OF THE MEANING OF RELEVANCE AND SUGGESTS EVALUATING SEVERITY IN TERMS OF INJUSTICE RATHER THAN INJURY. TABLES, FOOTNOTES, 57 REFERENCES, A QUESTIONNAIRE, AND THE THREE NEWSPAPER STORIES ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED)