Evidence-based treatments (EBTs) for youth are typically developed and established through studies in research settings designed to ensure treatment integrity, that is, protocol adherence and competence by therapists. An important question for implementation science is how well integrity is maintained when these EBTs are delivered in community settings. The present study investigated whether the integrity achieved by therapists in community settings achieved a benchmark set by therapists in a research setting when they delivered the same EBT-an individual cognitive-behavioral treatment (ICBT) for youth anxiety. Therapists (N = 29; 68.97% White; 13.79% male) provided ICBT to 68 youths (M age = 10.60 years, SD = 2.03; 82.35% White; 52.94% male) diagnosed with a principal anxiety disorder in research or community settings. Training and supervision protocols were the same across settings. Two independent teams of trained coders rated 744 sessions using observational instruments designed to assess ICBT adherence and competence. Both adherence and competence were higher in the research setting. Group differences in competence were consistent across treatment, but differences in adherence were most pronounced when treatment shifted to exposure, widely viewed as the most critical component of ICBT. When using the benchmarks from the research setting, therapists from the community settings fell short for indices of adherence and competence. However, given differences between therapists and clients, as well as the fact that treatment outcomes were similar across settings, our findings raise questions about whether it is appropriate to use treatment integrity benchmarks from research settings for community.