The study assessed the effects of different amounts of sleep restriction on slow wave sleep (SWS) in the ensuing recovery nights. After one adaptation night and an 8-hr baseline night, six healthy men were individually studied during and following five nights in which sleep was reduced to 5, 4, 3, 2, and 1 hr with a 1-week interval between conditions. Each sleep reduction was followed by an 8-hr recovery night. Finally, a second 8-hr baseline night was recorded. A trend analysis revealed that SWS amount in recovery nights increases with decreasing previous sleep duration. Regression analyses showed that, within each participant, the rebound of SWS after a sleep reduction is predicted better by the total duration of sleep than by the specific amount of SWS lost.