We examined bidirectional, temporal associations between daily sleep and affect under naturally restricted (school) and unrestricted (vacation) sleep opportunities, while incorporating valence (positive/negative) and arousal (high/low) dimensions of affect. Sleep and affect were measured over 2 weeks of school and 2 weeks of vacation in 205 adolescents (54.1% females, Mage = 16.9 years), providing 5,231 days of data. Total sleep time and sleep efficiency were measured using actigraphy and sleep diary. High‐ and low‐arousal positive and negative affect were self‐reported each afternoon. Between‐ and within‐person sleep–affect associations were tested using cross‐lagged, multilevel models. Lagged outcome, day of the week, study day and socio‐demographics were controlled. Bidirectional associations between self‐report sleep and affect were found between‐persons: longer self‐report total sleep time associated with lower high‐ and low‐arousal negative affect. Higher high‐arousal positive affect associated with longer actigraphy total sleep time between‐persons, but predicted shorter same‐night actigraphy total sleep time within‐persons. Results did not differ between school and vacation. Significant within‐person random effects demonstrate individual differences in daily sleep–affect associations. Associations differed based on sleep measurement and affect dimensions, highlighting the complex sleep–affect relationship. Strong between‐person associations between self‐report sleep and affect suggest improving either sleep or mood may benefit the other; alternatively, addressing a common cause may lead to changes in both sleep and affect. Although overall high‐arousal positive affect was protective of sleep duration, on a day‐to‐day basis, higher‐than‐usual high‐arousal positive affect may reduce sleep duration on nights it is experienced. Further research needs to identify causes of individual differences in sleep–affect associations.