For safety reasons as well as for experimental control, driver sleepiness experiments are often run in simulators. The aim of this study is to compare the development of driver sleepiness in an advanced driving simulator experiment with real road driving. Sixteen drivers participated in the experiment, which included daytime and night-time driving on a real motorway and in an advanced driving simulator. The results showed that there were significantly higher levels of Karolinska Sleepiness Scale (KSS) ratings, longer blink durations, lower percentage of gazes directed toward the road center (PRC), and higher speed, in the simulator compared to the real road. The pattern of change with task progression was consistent between simulator driving and real road driving for all investigated variables, that is, speed, KSS, line crossings, blink duration, and PRC. However, the relative differences were higher during daytime compared to night-time driving. Results from a questionnaire showed that it was more boring and more demanding to stay alert in the simulator. In conclusion, the development of sleepiness over time is similar in the simulator as compared to the real road, though the absolute sleepiness level is higher in the simulator.