IntroductionPoor sleep health is common among individuals in early treatment for substance use disorders (SUDs) and may serve an important role in predicting SUD outcomes. However, sleep parameters have been inconsistently linked with risk of relapse, perhaps because previous research has focused on mean values of sleep parameters (e.g., total sleep time [TST], sleep efficiency [SE], and sleep midpoint [SM]) across multiple nights rather than night-to-night fluctuations (i.e., intraindividual variability [IIV]). The current study assessed sleep across the first week of SUD treatment, with the aim of prospectively examining the relationship between mean and IIV of TST, SE, and SM and treatment completion and relapse within one-month post-treatment. MethodsTreatment-seeking adults (N = 23, Mage = 40.1, 39% female) wore an actigraph to assess sleep for one week at the beginning of an intensive outpatient program treatment. Electronic medical record and follow-up interviews were utilized to determine treatment outcomes. ResultsGreater IIV in TST was associated with higher odds of relapse (OR = 3.55, p =.028). Greater IIV in SM was associated with lower odds of treatment completion, but only when removing mean SM from the model (OR = 0.75, p =.046). DiscussionNight-to-night variability in actigraphy-measured TST is more strongly associated with SUD treatment outcomes than average sleep patterns across the week. Integrating circadian regulation into treatment efforts to improve SUD treatment outcomes may be warranted. Given the small sample size utilized in the present study, replication of these analyses with a larger sample is warranted.