Objective We aimed to study fatigue and sleep in registrars working 12-hour rotating shifts in our tertiary neonatal intensive unit. Methods and participants This study involved neonatal registrar’s working day (08:00–21:00) and night (20:30–08:30) shifts. Participants maintained a sleep diary, answered a self-reported sleepiness questionnaire assessing subjective sleepiness, and performed a 10-minute psychomotor vigilance task (PVT) at the start and end of each shift. Primary outcomes: (1) Fatigue at the (i) “start vs end” of day and night shifts, (ii) end of the “day vs night” shifts, and (iii) end of “first vs last shift” in block of day and night shifts. (2) Duration and quality of sleep before the “day vs night” shifts. Mean reaction time (RTM), relative coefficient of variation (RTCV), and lapses (reaction time > 500ms) were used as measures of fatigue on PVT. Secondary outcome: Subjective sleepiness (self-reported sleepiness questionnaire) at the ‘start vs end” of day and night shifts. Results Fifteen registrars completed the study. Acuity was comparable for all shifts. (1) Psychomotor responses were impaired at the end vs start of day shifts [RTM (p = 0.014), lapses (p = 0.001)], end vs start of night shifts [RTM (p = 0.007), RTCV (p = 0.003), lapses (p<0.001)] and end of night vs day shifts [RTM (p = 0.007), RTCV (p = 0.046), lapses (p = 0.001)]. Only lapses were significantly increased at the end of the last (p = 0.013) vs first shift (p = 0.009) in a block of day and night shifts. (2) Duration of sleep before the night (p = 0.019) and consecutive night shifts was decreased significantly (p = 0.034). Subjective sleepiness worsened after day (p = 0.014) and night shifts (p<0.001). Conclusion Fatigue worsened after the 12-hour day and night shifts with a greater change after night shifts. Lapses increased after block of day and night shifts. Sleep was decreased before night shifts. Our findings need to be confirmed in larger studies.