Aims: Neurobehavioural performance is typically worse during night shifts than dayshifts because the sleep/wake cycle and circadian rhythm are misaligned. Working multiple shorter shifts per day, to allow some sleep at night and some work during the daytime, may improve night-time performance. However, the effect of splitting the work-rest schedule on night-time performance is not currently clear. Therefore, the aim of this study was to compare performance during split and consolidated work-rest schedules at different times of the day. Methods: Twenty-nine male participants lived in a time-isolation laboratory for 12 days. Participants were scheduled to one of two 28-h forced desynchrony (FD) protocols. Each provided the same total time in bed (TIB), either consolidated (9.3h TIB/28h) or split (2 x 4.7h TIB/28h). Neurobehavioural performance was assessed with a psychomotor vigilance task (PVT), 2h following waking and every 2.5h thereafter. Response time (RT) relative to baseline was the measure of performance. A nadir at 0500h on the first day and a period of 24.2h were assumed to estimate circadian phase. Results: Participants in the consolidated (n=13) and split (n=16) conditions were similar in terms of age and BMI. Mixed effects model analyses indicated no overall difference between conditions in relative RT. However, there was a main effect of circadian phase and a circadian phase x condition interaction. Relative RT was fastest at the circadian acrophase and slowest around the circadian nadir. Relative RT during the split schedule was faster than the consolidated schedule around the circadian nadir. Discussion: Overall performance did not differ by splitting the work-rest schedule. However, in shortening the duration of wakefulness between sleep periods, performance during the night was maintained at a higher level than for the consolidated schedule. Therefore, this schedule may be beneficial to industries that require a high standard of performance sustained around the clock.