Background and Aims: University students with risky drinking are a clear target group for intervention via smartphone apps. This study compared three different apps over a 20-week period, for university students with hazardous and excessive drinking patterns. Materials and Methods: Students from six campuses were invited to a three-armed trial (A). Those with hazardous alcohol use (n = 2166) were randomly assigned to one of two smartphone apps offer- ing feedback on real-time estimated blood alcohol concentration (eBAC) levels, or to a control group, with three follow-ups at 6, 12 and 20  weeks. At 6  weeks, participants in the app groups with excessive weekly alcohol consumption of >9 (women) or >14 (men) drinks per week (n  =  257), were offered participation in a second trial (B); con-senters (n  =  186) were randomly assigned to a skills-based app or a waitlist group, and compared with an assessment-only control group. Results: Six-week analyses (n = 2166) replicated our earlier trial from 2014, re-confirming earlier results: the Promillekoll app was associated with higher quantity and frequency of drinking compared to controls, and a higher risk for excessive drinking; the PartyPlanner group did not differ from controls. Lower-risk drinkers from trial A (n = 1177) up to 20 weeks did not differ from controls on main outcomes. However, sub-analyses showed that individuals with higher consumption had higher motivation to reduce intake. In both intervention groups, con-sumption was lower for more highly motivated participants compared to controls at 6- and 20-week follow-ups. Latent class analysis of par- ticipants in both trials (n = 2166) revealed a class (n = 146) that drank several days a week and that differed significantly from the remain- ing cohort in gender, age, and alcohol consumption. For this class, access to the Promillekoll app appeared marginally associated with lower quantity over time; access to the skills-based TeleCoach app was clearly associated with fewer drinking days up to 20 weeks. Conclusions: Smartphone apps targeting eBAC can influence drink-ing levels up to 20 weeks for university students with hazardous use and higher motivation to reduce their drinking. A skills-based app that reduces intake among students with excessive weekly consumption can be particularly effective for students with daily drinking habits.