The 'sleep to forget and sleep to remember hypothesis' proposes that sleep weakens the emotional tone of an experience while preserving or even enhancing its content. Prior experimental research however shows contradictory findings on how emotional reactivity changes after a period of sleep, likely explained by methodological variations. By addressing these inconsistencies, we investigated the mitigating effect of overnight sleep on emotional reactivity triggered by memory reactivation. Using a karaoke paradigm, we recorded participants' singing of two songs, followed by exposing them to one of the recordings (Rec1) to induce an embarrassing episode. After a 12-h period of either day-time wakefulness (N=20) or including night-time sleep (N=20), we assessed emotional reactivity to the previously exposed recording (Rec1) and the newly exposed recording (Rec2). Emotional reactivity was assessed with a physiological measure of facial blushing as main outcome and subjective ratings of embarrassment and valence. Sleep and wake were monitored with diaries and actigraphy. The embarrassing episode was successfully induced as indicated by objective and subjective measures. After controlling for an order effect in stimulus presentation, we found a reduction in blushing response to the reactivated recording (Rec1) from pre- to post-sleep compared to wakefulness. However, emotional reactivity to the reactivated recording (Rec1) and the new recording (Rec2) did not differ after sleep and wakefulness. This study shows that facial blushing was reduced following overnight sleep, while subjective ratings were unaffected. Whether the beneficial effect of sleep is due to changes in memory representation or rather emotion regulation remains elusive.