Daily covariation of three sleep indicators (quantity, quality, postsleep restedness), anxiety symptoms, and caffeine and alcohol consumption was observed in 283 emerging adults (M _age = 19.9). Participants completed web-based sleep diary surveys every morning and evening for 7 consecutive days. Cross-lagged models suggested that sleep quality and restedness (reported each morning) predicted students’ anxiety levels (reported each evening) and that anxiety levels predicted all three sleep indicators. Furthermore, pre-bedtime caffeine consumption predicted worse sleep that night. Controlling for the effects of weekdays versus weekends produced virtually identical effects to the 7-day models, although weekend alcohol consumption emerged as a predictor of postsleep restedness. These data support bidirectionality between sleep and anxiety symptoms in emerging adults attending college/university and highlight potential avenues for intervention.