Although on-campus residence allows easier access to campus facilities, existing studies showed mixed results regarding the relationship between college residence and students' well-being indicators, such as sleep behaviors and mood. There was also a lack of studies investigating the role of chronotype in the relationship between on-campus residence and well-being. In particular, the temporal relationships among these factors were unclear. Hence, this longitudinal study aims to fill in these gaps by first reporting the well-being (measured in terms of mood, sleep, and quality of life) among students living on and off campus across two academic semesters. We explored factors predicting students' dropout in university residences. Although students living on campus differ in their chronotypes, activities in campus residence (if any) are mostly scheduled in the nighttime. We therefore tested if individual differences in chronotype interact with campus residence in affecting well-being. Our final sample consisted of 215 campus residents and 924 off-campus-living students from 10 different universities or colleges in Hong Kong or Macau. Their mean age was 20.2 years (SD=2.3); 6.5% of the participants are female. Participants completed self-reported questionnaires online on their sleep duration, sleep quality, chronotype, mood, and physical and psychological quality of life. Across two academic semesters, we assessed if students living on and off campus differed in our well-being measures after we partialed out the effects of demographic information (including age, sex, family income, and parents' education) and the well-being measures at baseline (T1). The results showed that, campus residents exhibited longer sleep duration, greater sleep efficiency, better sleep quality, and less feeling of stress than off-campus-living students. From one semester to the next, around 10% of campus residents did not continue to live on campus. Logistic regression showed that a morning type was the strongest factor predicting dropout from campus residence. Chronotype significantly moderated the effects of campus residence on participants' physical and psychological quality of life. Although morning-type off-campus-living students have better well-being than their evening-type peers living off campus, morning-type campus residents had worse well-being than other campus residents and they were more likely to discontinue living on campus after one semester. Our findings bear practical significance to college management that morning-type campus residents are shown to be experiencing deteriorating well-being. The authorities may need to review and revise the room-allocation policy in campus residence in improving the well-being among campus residents.