This study examined the effects of task and time-on-task on fatigue symptoms in overnight driving. Four participants drove an instrumented car 1200 km overnight and completed the same trip as passengers on another night. Subjective ratings of drowsiness, eye blink frequency and duration, microsleeps, and steering-wheel inputs were analysed as a function of time-on-task, and for separate samples when meeting oncoming heavy vehicles. Four video cameras were used to monitor the road view and the face of both the driver and passenger. In terms of eye closure duration, the reported microsleeps were shorter while driving (mean = 0.7 s, SD = 0.2 s) than as a passenger (mean = 2.6 s, SD = 2.0 s). Blink frequency increased with time-on-task as expected, indicating tiredness, and decreased when approaching an oncoming heavy vehicle, indicating attentive response to a potential critical situation. No consistent effect of time-on-task on high-frequency steering-wheel inputs when meeting oncoming heavy vehicles was found. The results raise the important question of what makes a driver wake from a microsleep earlier than a passenger and, given proper monitoring of long eyelid closures, what the proper intervention should be.