Abstract The “sleep to forget and sleep to remember” hypothesis states that sleep attenuates the emotional tone of a memory while strengthening its factual content. However, previous experimental research has yielded inconsistent results, associating sleep with the reduction, enhancement, or maintenance of the emotional tone of memories. Although the hypothesized process may necessitate multiple nights of sleep, most studies have relied on single-night protocols. To address this, we further investigated whether immediate sleep diminishes emotional reactivity triggered by memory reactivation after one week. In a karaoke paradigm, we recorded participants’ singing of two songs and played back one of their recordings (rec1) to induce an embarrassing episode either in the early afternoon (delayed sleep group; N = 25) or the evening (immediate sleep group; N = 25). One week later, we assessed participants’ emotional reactions to the re-exposed recording (rec1) and a newly introduced recording (rec2). Emotional reactivity was assessed using facial blushing as a primary physiological measure and subjective ratings of embarrassment, valence, and blushing. Sleep was monitored using diaries. While the embarrassing episode was successfully induced, Bayesian mixed-effects models revealed reduced facial blushing and more negative valence ratings from initial exposure to re-exposure (rec1) after both a shorter and longer interval to sleep. These changes were nonspecific to the reactivated recording (rec1) and were also observed for the new recording (rec2). Other subjective measures remained unchanged. This study demonstrates that neither the time interval to sleep following encoding nor memory reactivation influenced long-term emotional reactivity, leaving sleep’s role in emotional memory processing elusive.