Most night workers are unable to adjust their circadian rhythms to the atypical hours of sleep and wake. Between 10% and 30% of shiftworkers report symptoms of excessive sleepiness and/or insomnia consistent with a diagnosis of shift work disorder (SWD). Difficulties in attaining appropriate shifts in circadian phase, in response to night work, may explain why some individuals develop SWD. In the present study, it was hypothesized that disturbances of sleep and wakefulness in shiftworkers are related to the degree of mismatch between their endogenous circadian rhythms and the night-work schedule of sleep during the day and wake activities at night. Five asymptomatic night workers (ANWs) (3 females; [mean ± SD] age: 39.2 ± 12.5 yrs; mean yrs on shift = 9.3) and five night workers meeting diagnostic criteria (International Classification of Sleep Disorders [ICSD]-2) for SWD (3 females; age: 35.6 ± 8.6 yrs; mean years on shift = 8.4) participated. All participants were admitted to the sleep center at 16:00 h, where they stayed in a dim light (<10 lux) private room for the study period of 25 consecutive hours. Saliva samples for melatonin assessment were collected at 30-min intervals. Circadian phase was determined from circadian rhythms of salivary melatonin onset (dim light melatonin onset, DLMO) calculated for each individual melatonin profile. Objective sleepiness was assessed using the multiple sleep latency test (MSLT; 13 trials, 2-h intervals starting at 17:00 h). A Mann-Whitney U test was used for evaluation of differences between groups. The DLMO in ANW group was 04:42 ± 3.25 h, whereas in the SWD group it was 20:42 ± 2.21 h (z = 2.4; p < .05). Sleep did not differ between groups, except the SWD group showed an earlier bedtime on off days from work relative to that in ANW group. The MSLT corresponding to night work time (01:00-09:00 h) was significantly shorter (3.6 ± .90 min: [M ± SEM]) in the SWD group compared with that in ANW group (6.8 ± .93 min). DLMO was significantly correlated with insomnia severity (r = -.68; p < .03), indicating that the workers with more severe insomnia symptoms had an earlier timing of DLMO. Finally, SWD subjects were exposed to more morning light (between 05:00 and 11:00 h) as than ANW ones (798 vs. 180 lux [M ± SD], respectively z = -1.7; p < .05). These data provide evidence of an internal physiological delay of the circadian pacemaker in asymptomatic night-shift workers. In contrast, individuals with SWD maintain a circadian phase position similar to day workers, leading to a mismatch/conflict between their endogenous rhythms and their sleep-wake schedule.