Prior studies have established an association between sleep problems during early adolescence and heavy alcohol use/alcohol use disorder (AUD) risk in late adolescence. Less research has explored the association between sleep problems and heavy alcohol use during young adulthood, the period when AUD onset peaks. Moreover, research to date has primarily utilized cross-sectional, between-subjects’ methods to examine this relationship, with limited focus on the potential intraindividual variation in these behaviors. Multilevel modeling techniques are well-suited to examine the variability in sleep problems and risky alcohol use over time and the dynamic bidirectional relations among these behaviors. This article reports on 42 heavy-drinking college students at-risk for an AUD based on their responses to a validated alcohol screener who completed daily diaries of sleep and alcohol use and wore a sleep-wake activity monitor (i.e., Philips Respironics Actiwatch 2™) daily for 7 days yielding a total of 294 reports. Hierarchical linear models demonstrated that days of heavy drinking predicted delayed bed and wake times within individuals and those individuals who tended to drink more heavily on average had shorter sleep durations. Conversely, days of shorter sleep duration, earlier wake times, and greater perceived sleep quality upon waking predicted greater alcohol use within individuals, and those who tended to feel more alert upon waking drank more on average. These results highlight important within- and between-person variability in the associations among objective and subjective sleep-related problems and at-risk drinking among young adults. Further, the results have implications for alcohol prevention/intervention strategies for young adults at risk for AUDs.