We study motives for executive stock option backdating, the practice of changing the grant dates of current options to dates in the past using hindsight. We find that smaller, younger and less profitable firms tend to be more heavily involved in backdating. These results are consistent with the retention hypothesis. In line with the incentive hypothesis, we find that backdating occurs more for options that are out-of-the-money. We derive some evidence for the agency hypothesis, in the sense that backdating companies have a larger percentage of inside directors. However, contrary to this hypothesis, we conclude that backdating firms have better protection for minority shareholders compared to firms that do not backdate.