Blockade of the inflammatory cytokine tumor necrosis factor (TNF) in depressed patients with increased inflammation has been associated with decreased depressive symptoms. Nevertheless, the impact of TNF blockade on sleep in depressed patients has not been examined. Accordingly, sleep parameters were measured using polysomnography in 36 patients with treatment resistant major depression at baseline and 2weeks after 3 infusions (week 8) of either the TNF antagonist infliximab (n=19) or placebo (n=17). Markers of inflammation including c-reactive protein (CRP) and TNF and its soluble receptors were also assessed along with depression measured by the 17-item Hamilton Depression Rating Scale. No differences in sleep parameters were found as a function of infliximab treatment over time. Nevertheless, wake after sleep onset (WASO), the spontaneous arousal index and sleep period time significantly decreased, and sleep efficiency significantly increased, from baseline to week 8 in infliximab-treated patients with high (CRP>5mg/L) (n=9) versus low inflammation (CRP⩽5mg/L) (n=10), controlling for changes in scores of depression. Stage 2 sleep also significantly decreased in infliximab-treated patients with high versus low inflammation. Decreases in soluble TNF receptor 1 (sTNFR1) significantly correlated with decreases in WASO and increases in sleep efficiency in infliximab-treated subjects with high inflammation. Placebo-treated subjects exhibited no sleep changes as a function of inflammation, and no correlations between inflammatory markers and sleep parameters in placebo-treated patients were found. These data suggest that inhibition of inflammation may be a viable strategy to improve sleep alterations in patients with depression and other disorders associated with increased inflammation.