Laboratory-based sleep manipulations show asymmetries between positive and negative affect, but say little about how more specific moods might change. We report extensive analyzes of items from the Positive and Negative Affect Scale (PANAS) during days following nights of chronic sleep restriction (6 h sleep opportunity), during 40 h of acute sleep deprivation under constant routine conditions, and during a week-long forced desynchrony protocol in which participants lived on a 28-h day. Living in the laboratory resulted in medium effects sizes on all positive moods (Attentiveness, General Positive Affect, Joviality, Assuredness), with a general deterioration as the days wore on. These effects were not found with negative moods. Sleep restriction reduced some positive moods, particularly Attentiveness (also General Positive), and increased Hostility. A burden of chronic sleep loss also led to lower positive moods when participants confronted the acute sleep loss challenge, and all positive moods, as well as Fearfulness, General Negative Affect and Hostility were affected. Sleeping at atypical circadian phases resulted in mood changes: all positive moods reduced, Hostility and General Negative Affect increased. Deteriorations increased the further participants slept from their typical nocturnal sleep. In most cases the changes induced by chronic or acute sleep loss or mistimed sleep waxed or waned across the waking day, with linear or various non-linear trends best fitting these time-awake-based changes. While extended laboratory stays do not emulate the fluctuating emotional demands of everyday living, these findings demonstrate that even in controlled settings mood changes systematically as sleep is shortened or mistimed.