This study examined self-reported academic dishonesty at a midsize public university. Students (N = 492) rated the likelihood they would cheat after accepting to abide by each of eight honor code pledges before Internet-based assignments and examinations. The statements were derived from honor pledges used by different universities across the United States and varied in length, formality, and the extent to which the statements included consequences for academic dishonesty. Longer, formal honor codes with consequences were associated with a lower likelihood (...) to cheat. Results showed a significant three-way interaction and suggest how to best design honor codes. (shrink)