Five decades of randomized trials research have produced dozens of evidence-based psychotherapies (EBPs) for youths. The EBPs produce respectable effects in traditional efficacy trials, but the effects shrink markedly when EBPs are tested in practice contexts with clinically referred youths and compared to usual clinical care. We considered why this might be the case. We examined relevant research literature and drew examples from our own research in practice settings. One reason for the falloff in EBP effects may be that so little youth treatment research has been done in the context of everyday practice. Researchers may have missed opportunities to learn how to make EBPs work well in the actual youth mental health ecosystem, in which so many real-world factors are at play that cannot be controlled experimentally. We sketch components and characteristics of that ecosystem, including clinically referred youths, their caregivers and families, the practitioners who provide their care, the organizations within which care is provided, the network of youth service systems (e.g., child welfare, education), and the policy context (e.g., reimbursement regulations and incentives). We suggest six strategies for future research on EBPs within the youth mental health ecosystem, including reliance on the deployment-focused model of development and testing, testing the mettle of current EBPs in everyday practice contexts, using the heuristic potential of usual care, testing restructured and integrative adaptations of EBPs, studying the use of treatment response feedback to guide clinical care, and testing models of the relation between policy change and EBP implementation. (Contains 2 figures, 2 tables, and 1 footnote.)