Sustained flight operations are likely to produce fatigue and performance decrement in aviators. We assessed changes in cognitive performance using a work/rest schedule modeled on successive long-range attack missions. Twelve subjects performed several subtests of the Unified Tri-Service Cognitive Performance Assessment Battery and the Walter Reed Performance Assessment Battery 18 times during a simulated sustained operation. The scenario consisted of a 9-hr planning session followed by a 4-hr rest period and a 14-hr daytime mission. After 6 hr of rest, subjects repeated this schedule with a nighttime mission. For two spatial tests, subjects showed linear increases in response rate and one of its components, error rate. Subjects appeared to change strategy as the study progressed, possibly exchanging a higher failure rate for a savings in time. Any tendency to take chances when fatigued may have serious implications for aircrew in sustained operations.