A critical task for psychotherapy research is to create treatments that can be used by community clinicians. Streamlining of psychotherapies is a necessary first step for this purpose. We suggest that neurobiological knowledge has reached the point of providing biologically meaningful behavioral targets, thus guiding the development of effective, simplified psychotherapies. This view is supported by the Research Domain Criteria (RDoC) Project, which reflects the field's consensus and recognizes the readiness of neurobiology to guide research in treatment development. 'Engage' is an example of such a streamlined therapy. It targets behavioral domains of late-life depression grounded on RDoC constructs using efficacious behavioral strategies selected for their simplicity. 'Reward exposure' targeting the behavioral expression of positive valence systems' dysfunction is the principal therapeutic vehicle of 'Engage'. Its first three sessions consist of direct 'reward exposure', but the therapists search for barriers in three behavioral domains, that is, 'negativity bias' (negative valence), 'apathy' (arousal) and 'emotional dysregulation' (cognitive control), and add strategies targeting these domains when needed. The end result is a structured, stepped approach using neurobiological constructs as targets and as a guide to personalization. We argue that the 'reduction' process needed in order to arrive to simplified effective therapies can be achieved in three steps: (1) identify RDoC constructs driving the syndrome's psychopathology; (2) create a structured intervention utilizing behavioral and ecosystem modification techniques targeting behaviors related to these constructs; (3) examine whether the efficacy of the new intervention is mediated by change in behaviors related to the targeted RDoC constructs.