Objective: As alcohol is often consumed for social purposes, we aimed to explore how restrictions during the first Danish COVID-19 lockdown affected the alcohol use among adolescents aged 15–20. Method: In May 2020, 11,596 15- to 20-year-olds from two subpopulations answered a survey regarding their alcohol use and social life, as well as changes to these, during the Danish lockdown. Using survey data from all participants, we performed a multinomial logistic regression to assess the association between determinants of alcohol use and perceived change in alcohol use during the Danish lockdown. We used longitudinal data from one subpopulation ( n=1869) to perform negative binomial regressions exploring changes in frequency of alcohol use from 2019 to 2020. Results: Of all participants, 59% drank less, 75% had fewer in-person social interactions and 56% met more frequently online during lockdown. Girls were more likely than boys to have a perceived decrease in alcohol use during lockdown (odds ratio (OR)=1.41; 95% confidence interval (CI) 1.27–1.56). A perceived decrease in in-person social interaction during lockdown was associated with less drinking (OR=2.27; 95% CI 1.98–2.61), while a perceived increase in in-person social interaction during lockdown was associated with more drinking (OR=1.42; 95% CI 1.11–1.82) compared to unchanged drinking behaviour and social interaction. Conclusions: Adolescents in Denmark drank less during the Danish lockdown than before. Findings indicate that there is a close relationship between in-person social interactions and frequency of drinking. Drinking episodes when meeting online were rare and were not unambiguously associated with changes in drinking during lockdown.