The weekend is usually seen as a window of recovery. Thus, sleep before a day off may be less impaired than that before a workday. However, very few polysomnographical studies have investigated this hypothesis. Therefore, the aim of the present study was to compare sleep before a workday with that before a weekend. Seventeen teachers participated. Sleep was recorded with polysomnography on one weekday night during the workweek, and on a workday (Friday) followed by a day off. Sleep diaries and actigraphs were also used. Weekend sleep showed delayed bedtime and time of rising, a longer total sleep time (45 min), increased N3 and N1, and decreased N2 and REM. Sleep spindles were reduced. The results remained after truncation to the shortest common sleep duration (5 h). The increase in N3 from weekday sleep to Friday night sleep was positively correlated with N1 change (r = 0.853, P <= 0.001), and negatively correlated with N2 change (r = -0.614, P <= 0.001). Subjective ratings showed that weekend sleep was associated with less awakening problems and lower subjective arousal during the day. The authors concluded that weekend sleep was longer, and showed increased N3 and N1. The authors suggest that the N3 increase before the day off is a result of lower stress, while the N1 increase may be an effect of sleep spindle suppression via the increase of N3 (which would suppress sleep spindles), thus reducing N2 and enhancing N1.