<p>Self-reported sleep has shown a positive relationship with task performance at work. This study expands the understanding of the sleep-task performance relationship by exploring physiological aspects of sleep, more specifically sleep stages, and their impact on daily task performance. Specifically, we investigated intra-individual variations in sleep stages and their effects. We employed a three-week experience sampling method based on a sample of call centre employees (<i>n</i> = 26 individuals) recorded nightly for physiological measurements of rapid eye movement (REM) sleep, deep sleep/NREM sleep, total sleep time, and daily task performance (total range of observations 246 ≤ <i>n</i> ≤ 249). Sleep stages and duration were recorded with Somnofy^®, a noninvasive radar technology previously validated against polysomnography. Daily objective task performance metrics, including number of phone calls and total customer talk time, were derived from organizational records. Multilevel modelling showed a significant positive relationship between time spent in REM sleep and task performance (number of phone calls and talk time with customers) on the following day. Time spent in deep NREM sleep (stage N3) was not significantly related to task performance, while total sleep time was positively related to the number of calls made the following day. Analyses of the reversed relationships showed that task performance generally did not predict sleep outcomes the subsequent night, although greater customer talk time was linked to a reduction in REM sleep the subsequent night.</p>