Long-haul flight crews often operate a succession of transmeridian flights with approximately 24-hour rest periods (layovers) between flights. This creates an erratic pattern of environmental time cues to the circadian clock. Twenty-nine long-haul crew members (average age 52 years) were monitored before, during, and after four international trip patterns lasting 5-9 days, with flights crossing up to 8 time zones. Rectal temperature was recorded every 2 min. Sleep quantity and quality, and nap timing, were noted in a logbook. Significant periodicities in trip temperature data were detected by linear-non-linear least squares iterative multiple regression (1). Cycle-by-cycle temperature minima were estimated by multiple complex demodulation of each subject’s data (unmasked by adding 0.28°C to the raw temperature data whenever he reported being asleep; ref. 2). The average duty/rest cycle lasted 34.6 h (9.8 h of duty and a 24.8 h layover). Layover sleep episodes averaged 105 min shorter than pretrip sleep episodes. However, in 2/3 of layovers crew members slept twice, so that their total sleep per 24 h on trips averaged 49 min less than pretrip. Three main factors influenced layover sleep: 1) local time; 2) prior flight direction; and 3) the circadian cycle. When asked about their layover strategies, most crew members indicated that they tried to adapt to local time, but were only moderately sucessful. Figure 1 shows a preference for sleeping during local night. However, after eastward night flights (average off-duty time 1100 local time), 2/3 of crew members slept for several hours in the afternoon, and then again during local night. In contrast, after westward flights (average off-duty time 1400), crew members usually deferred sleep until local night, and this was the main layover sleep episode (80% of cases). After flights crossing fewer than 4 time zones, no one layover sleep pattern predominated (Figure 2). Eighty-two percent of subjects retained significant circadian variation in temperature during trips (average period 25.7 h). The average time of sleep onset was 2 min after the temperature minimum, and of wakeup 6.4 h after the temperature minimum (Figure 3), comparable to the circadian distributions of sleep onset and wakeup of desynchronized subjects in time-free environments (3). Two types of sleep episodes did not conform to this pattern. Afternoon sleep episodes after eastward night flights were broadly distributed in the circadian cycle, and were probably a response to acute sleep loss rather than to circadian physiology. Short sleep episodes towards the end of layovers after westward flights typically ended as temperature was falling. In this case, crew members had to wake up to go back on duty. These findings highlight the complex combination of factors that influence layover sleep during long-haul flight operations, when the circadian clock cannot follow the erratic environmental time cues. Operational constraints sometimes leave minimal room for flexibility in the timing of layovers. Long-haul crew members need effective countermeasure strategies to minimize their sleep loss during these operations (4).