STUDY OBJECTIVE: To evaluate the effect of experimental sleep fragmentation (sleep interruption; SI) on complex learning in an intradimensional-extradimensional (ID/ED) set-shifting task in rats. DESIGN: A sleep fragmentation paradigm of intermittent forced locomotion was validated in adult rats by examining electrographic effects. Discrimination task performances were assessed in rats following sleep fragmentation or 2 control conditions. PARTICIPANTS: 41 young adult male Fischer-Norway rats. INTERVENTION: A treadmill was used to produce 30 awakenings/h for the 24-h period prior to testing. Exercise control rats received an equivalent amount of treadmill-induced locomotion that permitted 30-minute pauses to allow consolidated sleep. MEASUREMENT AND RESULTS: SI rats were selectively impaired on the extradimensional-shift phase of the task, taking significantly more trials to achieve criterion performance (15.4 +/- 2.0) than either control group (cage control = 10.4 +/- 0.9; exercise control = 6.3 +/- 0.2). The SI schedule reduced the average duration of nonREM sleep (NREMS) episodes to 56 s (baseline = 182 s), while the exercise control group increased average NREMS episode duration to 223 s. Total (24-h) NREMS time declined from 50% during baseline to 33% during SI, whereas rapid eye movement sleep (REMS) was absent in SI animals (7% during baseline and 0% during SI), and time spent awake increased proportionally (from 43% during baseline to 67% during SI). CONCLUSION: 24-hour SI produced impairment in an attentional set-shifting that is comparable to the executive function and cognitive deficits observed in humans with sleep apnea or after a night of experimental sleep fragmentation.