Digital phenotyping extends the assessment of alcohol use disorder by incorporating data collected between clinical visits in the patient’s everyday environment. It includes brief, repeated self-reports as well as passive data obtained from smartphones and other devices, interpreted in conjunction with information on mood, alcohol craving, sleep, activity, and environmental context. The aim of this paper is to discuss the significance of this approach in identifying increasing relapse risk and in guiding the selection of appropriate therapeutic interventions. At present, the principal clinical value of digital phenotyping lies in its potential to enable earlier detection of patient deterioration and the more rapid implementation of an adequate therapeutic response. Its main limitations include methodological heterogeneity, missing data, privacy-related concerns, and the need for further clinical validation.