<jats:sec><jats:title>Background</jats:title><jats:p>It is still unclear which biological changes are needed to recover from a major depressive episode. Current perspectives focus on cortical synaptic neuroplasticity. Measures of cortical responses evoked by transcranial magnetic stimulation (TMS) change with sleep homeostasic pressure in humans and approximate measures of synaptic strength in animal models. Using repeated total sleep deprivation as a model of antidepressant treatment, we aimed to correlate recovery from depression with these measures of cortical excitability.</jats:p></jats:sec><jats:sec><jats:title>Methods</jats:title><jats:p>We recorded electroencephalographic responses to TMS in the prefrontal cortex of 21 depressed inpatients with bipolar disorder treated with repeated sleep deprivation combined with light therapy. We performed seven TMS/electroencephalography sessions during one week and calculated three measures of cortical excitability.</jats:p></jats:sec><jats:sec><jats:title>Results</jats:title><jats:p>Cortical excitability progressively increased during the antidepressant treatment and as a function of time awake. Higher values differentiated responders from non‐responders at baseline and during and after treatment on all measures.</jats:p></jats:sec><jats:sec><jats:title>Conclusions</jats:title><jats:p>Changes in measures of cortical excitability parallel and predict antidepressant response to combined sleep deprivation and light therapy. Data suggest that promoting cortical plasticity in bipolar depression could be a major effect of successful antidepressant treatments, and that patients not responding could suffer a persistent impairment in their neuroplasticity mechanisms.</jats:p></jats:sec>