The purpose was to examine undergraduates' perceptions of the nature, frequency, and possible justification of dishonest behaviors in relation to ethnicity (African-American, White American) and religious involvement (low, medium and high);Fifteen dishonest behaviors were presented. Those rated "definitely cheating" were: taking an exam for another student, copying from someone's exam, giving or receiving signals and/or answers during an exam, using unauthorized exam notes, turning in a commercially prepared paper or one written by another student;Behaviors considered "probably cheating" were: bringing information to an exam on a disc, arranging seating in advance to facilitate copying, turning in another student's computer program or illegally accessing a program, copying from a source without footnoting, adding unused items to a bibliography, getting answers to previously administered exams, and unauthorized collaboration on assignments;Perceptions of dishonesty did not differ by ethnicity. However, students who participated less in religious activities were less likely than more involved students to consider the following behaviors to be cheating: copying without footnoting and unauthorized collaboration;When frequency of dishonest behaviors was examined, at least 50 to 60 percent of students overall indicated that the following actions occurred "a great deal" or "a fair amount": unauthorized collaboration, getting answers to previously administered exams, and copying without footnoting. Items categorized as infrequent were: taking an exam for another student, copying from someone's exam, and giving or receiving signals and/or answers during an exam. Few differences were found when comparisons were made based upon ethnicity and religious involvement;Students overall agreed that cheating was never justified under any circumstances. African-Americans were slightly but significantly more likely to agree that cheating is sometimes justified to pass a course, to stay in school, to receive a better grade, to pass a course for graduation, to keep a scholarship, or when a friend asks for help. Finally, when compared to most other students in the study, African-Americans students who were least involved in religious activities were not as likely to agree that cheating is never justified under any circumstances. Results are discussed in terms of previous findings reported in the literature.