Purpose Treatment engagement poses challenges for youth mental health providers. With the expansion of evidence-based treatments (EBTs), providers face complex decisions regarding how to engage youth and families using available information sources. This study investigated how EBT protocols are associated with the selection and delivery of engagement practices. Method Twenty engagement practices were coded in a sample of digital recordings of early treatment sessions (N = 193) from the Child STEPs in California study, a randomized trial testing modular treatment and community-implemented treatment for youth mental health problems. Data were collected on which protocols mental health providers reportedly used to guide their sessions and the protocols in which they had received training. We examined which information sources (i.e., the guiding protocol, other protocols in training history, unspecified source) were associated with observed engagement practices. Results In sessions guided by a protocol, most observed engagement practices were accounted for by the guiding protocol (p < .001), rather than protocols in training history or unspecified sources (p < .001). In sessions not guided by a protocol, most observed practices were accounted for by training history (p < .001). Practice frequency and extensiveness was generally greater when a protocol guided the session. Conclusions Inclusion in protocols is associated with the selection and delivery of engagement practices, but this strategy might be insufficient for supporting the use of the full range of engagement practices supported by evidence. Supports are needed that leverage the engagement evidence base to ensure that selected practices empirically fit the engagement needs of youth and families.