International dissemination of evidence-based interventions calls for rigorous evaluation. As part of an evaluation of critical time intervention (CTI) for homeless people and abused women leaving Dutch shelters, this study assessed fidelity in two service delivery systems and explored factors influencing model adherence. Data collection entailed chart review (n = 70) and two focus groups with CTI workers (n = 11). The intervention obtained an overall score of three out of five (fairly implemented) for compliance fidelity and chart quality combined. Fidelity did not differ significantly between service systems, supporting its suitability for a range of populations. The eight themes that emerged from the focus groups as affecting model adherence provide guidance for future implementation efforts.