Thrilled to share a new consensus statement just published in Nature Mental Health: "Ambulatory assessment in mental health: expert consensus and recommendations" Ambulatory assessment — repeatedly capturing people's thoughts, feelings, and behaviors in real time as they go about daily life — has proliferated across mental health science. That growth is exciting, but it has also produced enormous methodological heterogeneity, which makes studies hard to compare and standards hard to pin down. To address these issues, 26 of us spanning psychology, psychiatry, computer science, and industry worked through a Delphi-adapted process to reach consensus on best practices across eight areas: sampling design, timing and duration, psychometrics and measurement standards, multimodal data, compliance, data analysis, open science and ethics, and clinical applications. A few themes run throughout. Align your design, measurement, and analysis with the actual research question rather than defaulting to convention. Take the temporal dynamics of your constructs seriously (e.g., the right sampling frequency for momentary affect is not the right one for rare events like self-injury). And build interdisciplinary collaboration and participant safety into the work from the outset, not as afterthoughts. Our aim is for this to be a practical resource, especially for those newer to these methods, for conducting ambulatory assessment with greater rigor, and for moving toward the dynamic, personalized understanding of mental health these tools promise. Glad to have contributed to this group effort. Open-access read cube link here: rdcu.be/fnxDZ