The study consisted of an on-road validation of a previous study of factors affecting driver alertness which used a driving simulator to investigate the effects of driving time, acoustic noise, and task complexity on driver performance. The on-road study closely duplicated the low-event, long-duration driving environment. In general, the findings of the simulator study were verified, with exceptions attributable to the inherent lack of experimental control in the field-test situations. It was confirmed that lateral road position error significantly increased as a function of time, as also does the occurrence of the alpha rhythm in the electroencephalogram (EEC). The number of small (2 degrees) steering wheel reversals also tended to decrease as time progressed. Although not demonstrated in the simulator study, the use of an automatic speed controller was found to foster decreases in alertness, as evidenced by changes in heart rate and the theta EEG component. Loud continuous noise acted as a stressful stimulus -- physiological measures indicated high arousal, while a performance measure showed higher error scores than shown at lower noise levels. Many of the measures were found to supplement each other, leading to a recommendation for a multiple regression analysis to develop an advanced index of driver alertness.