Purpose of the Review <p>The conventional approaches for diagnosing and monitoring major depressive disorder such as clinical interviews, structured assessments, and self-reporting scales. Moreover, it relies heavily on patient self-disclosure, clinician judgment, and observational data, which can lead to subjectivity and recall bias. These approaches are unable to capture the real-time dynamics of symptoms. This review aims to synthesise the current state of digital biomarkers, critically examine their clinical relevance, real-world applicability and alignment with DSM-5-TR symptoms and RDoC constructs.</p> Recent Findings <p>In our review, we found that passive digital biomarkers, including activity and mobility patterns, location diversity, sleep–wake variables, speech and language features, communication behaviour, and cognitive–motor posters, had meaningful correlations with core domains of depression symptoms such as anhedonia, psychomotor disturbance, sleep dysregulation, cognitive impairment and social withdrawal.</p> Summary <p>This review highlights the importance of combining DSM-5-TR symptomatology with RDoC psychological systems to improve the interpretability, standardisation, and translatability of digital phenotyping into clinical practice. Moving forward, digital psychiatry will need to focus on construct-level alignment, multimodal integration, and implementation in consultation, taking into account the feedback of clinicians, to assist in delivering an individual, sensitive, and holistic way of taking care of people with depression in various healthcare contexts.</p>