A naturalistic sample of N = 17 trainee therapists treating N = 30 patients in a psychodynamic doctoral training program was evaluated from a discovery-oriented perspective, inspired by the task analytic work of Jeremy D. Safran. A number of research results support the efficacy of integrative treatments, such as Safran’s metatherapeutic approach to working with patient—therapist intersubjective experience (Safran & Muran, 2000), highlighting therapist responsiveness to patient needs rather than strict adherence to manualized protocols as predictors of overall outcome (e.g., Katz et al., 2019; Owen & Hilsenroth, 2014). This study of early treatment sessions tested competing hypotheses that the trainee therapists would either demonstrate therapeutic processes consistent with the program’s psychodynamic training model or evidence of the “smuggling hypothesis” (Ablon & Jones, 1998), a psychodynamic therapist’s tendency to borrow prototypical cognitive—behavioral processes and interventions found in a naturalistic sample of experienced self-identified psychodynamic clinicians (Ablon, Levy, & Katzenstein, 2006). Results using the Psychotherapy Process Q-Set (Jones, 2000) suggested that treatment-as-usual for the sample of trainees was more consistent with a cognitive—behavioral therapy ideal prototype, supporting the latter prediction. Trainee therapists also demonstrated the ability to help their patients reduce self-reported symptoms within the first 3 months of psychotherapy and alliance significantly predicated reliable change, consistent with the broader psychotherapy research literature. Implications of the findings for psychotherapy training are discussed. (PsycINFO Database Record (c) 2019 APA, all rights reserved)