A primary community prevention approach in Iceland was associated with strong reductions of substance use in adolescents. Two years into the implementation of this prevention model in Chile, the aim of this study was to assess changes in the prevalence of adolescent alcohol and cannabis use and to discuss the impact of the COVID-19 pandemic on the substance use outcomes. In 2018, six municipalities in Greater Santiago, Chile, implemented the Icelandic prevention model, including structured assessments of prevalence and risk factors of substance use in tenth grade high school students every 2 years. The survey allows municipalities and schools to work on prevention with prevalence data from their own community. The survey was modified from an on-site paper format in 2018 to an on-line digital format in a shortened version in 2020. Comparisons between the cross-sectional surveys in the years 2018 and 2020 were performed with multilevel logistic regressions. Totally, 7538 participants were surveyed in 2018 and 5528 in 2020, nested in 125 schools from the six municipalities. Lifetime alcohol use decreased from 79.8% in 2018 to 70.0% in 2020 (X2 = 139.3, p < 0.01), past-month alcohol use decreased from 45.5 to 33.4% (X2 = 171.2, p < 0.01), and lifetime cannabis use decrease from 27.9 to 18.8% (X2 = 127.4, p < 0.01). Several risk factors improved between 2018 and 2020: staying out of home after 10 p.m. (X2 = 105.6, p < 0.01), alcohol use in friends (X2 = 31.8, p < 0.01), drunkenness in friends (X2 = 251.4, p < 0.01), and cannabis use in friends (X2 = 217.7, p < 0.01). However, other factors deteriorated in 2020: perceived parenting (X2 = 63.8, p < 0.01), depression and anxiety symptoms (X2 = 23.5, p < 0.01), and low parental rejection of alcohol use (X2 = 24.9, p < 0.01). The interaction between alcohol use in friends and year was significant for lifetime alcohol use (β = 0.29, p < 0.01) and past-month alcohol use (β = 0.24, p < 0.01), and the interaction between depression and anxiety symptoms and year was significant for lifetime alcohol use (β = 0.34, p < 0.01), past-month alcohol use (β = 0.33, p < 0.01), and lifetime cannabis use (β = 0.26, p = 0.016). The decrease of substance use prevalence in adolescents was attributable at least in part to a reduction of alcohol use in friends. This could be related to social distancing policies, curfews, and homeschooling during the pandemic in Chile that implied less physical interactions between adolescents. The increase of depression and anxiety symptoms may also be related to the COVID-19 pandemic. The factors rather attributable to the prevention intervention did not show substantial changes (i.e., sports activities, parenting, and extracurricular activities).