The study examined the effects of the amount of preceding sleep and work pace on sleepiness and cognitive performance during a 12-h dayshift. Twelve process operators (aged 28-56 years) completed a study with four single 12-h dayshifts and preceding night sleep in the laboratory. A simulated distillation process served as a work task. The 12-h shifts differed from each other in terms of the amount of preceding night sleep (23:00-06:30 hours or 2:30-6:30 hours) and work pace (slow or fast). All shifts contained four work simulation sessions of 1.5 h, and each of them included a 15-min alarm session. Cognitive performance was also measured with a 10-choice reaction time test and a mental subtraction test. Objective sleepiness was measured with a continuous electroencephalography/electro-oculography (EEG/EOG) recording during the work periods and with a sleep latency test. Subjective sleepiness at work was measured with the Karolinska Sleepiness Scale. Sleep debt increased the proportion of EEG/EOG-defined and subjective sleepiness at work, but did not impair work or test performance. The fatiguing effect of monotonous work as indicated by EEG/EOG-defined sleepiness was comparable with the effect of sleep debt. The alarm epochs in the middle of monotonous work temporarily decreased EEG/EOG-defined sleepiness. Sleep debt or monotonous work did not have a significant effect on the results of the sleep latency test. None of the sleepiness or performance measures indicated the impairment of a subject's functional capacity at the end of the 12-h shift. Our results suggest that monotonous work is at least as harmful as moderate sleep debt for alertness at work. The results support the view that the last hours of a single 12-h dayshift with frequent pauses are not associated with an increase in sleepiness or performance errors.