The software development industry often brings in security at the eleventh hour, right before developers throw the code over the wall--that is, deploy it into production--and ask, "Well, is it secure?" At this point, hilarity--for the objective observers, anyhow--ensues as security personnel work feverishly to shove crypto, firewalls, and all the other mechanisms at their disposal into the most egregious risk areas. This article describes how to create useful misuse cases within the development process, thus avoiding the headaches that can arise toward the end of the cycle.