An empirically based mathematical algorithm that predicts operator fatigue was developed for discrete event computer simulations tools. The algorithm estimates fatigue resulting from extended duty days and fragmented or reduced sleep. The algorithm would be useful in many human system integration problems such as predicting which crew duty schedules would produce the least fatigued operators or which crews would be most rested to complete a difficult assignment. It might also be used to degrade computer-generated forces from fatigue to more realistically estimate their effectiveness. During a Micro Saint simulation of two weeks of maritime activity, fragmented and reduced sleep produced by the contemporary US Navy watch schedules resulted in unacceptable fatigue levels over another crew duty schedule.