Summary Home visiting programs for new mothers have a broad range of goals, including improvements in maternal and child health, reductions in child maltreatment, and improvements in child development. Over 286,000 families facing adversity are served annually by 19 evidence-based models throughout the United States, which provide support, education, and connections to other services. Before 2020, only a small portion of visits were delivered through tele home visiting (teleHV; phone or video encounters). The Coronavirus Disease-2019 (COVID-19) pandemic required rapid adoption of teleHV; by early April 2020, 99% of home visits were by teleHV. Prior to 2020, teleHV was recommended only as a partial replacement for in-person visits and only after a relationship between the home visitor and family was established, because of the importance of the therapeutic relationship. This relationship is a key element across home visiting programs, because it serves as the foundation for teaching, mentoring, and collaborating with the family. There is limited evidence regarding teleHV, although we anticipate many families will use a mix of in-person visits and teleHV after COVID-19 restrictions are lifted due to greater flexibility and other advantages. We propose a secondary data analysis to make use of this natural experiment and determine if in-person visits during the establishment of the home visitor-family relationship is associated with better outcomes than teleHV during the establishment of this relationship. We will also explore if family characteristics associated with better outcomes vary between in- person home visiting and teleHV, which will contribute to our understanding of both acceptance and effectiveness of teleHV by family characteristics. We will consider both health outcomes (maternal depressive symptoms, breastfeeding, intimate partner violence [IPV]) and process outcomes (retention, visit attendance, screening completion). We will obtain data from a large, evidence-based home visiting program, Nurse-Family Partnership (NFP). NFP has approximately 260 sites across the United States; 1000 to 1600 pregnant first- time mothers enroll per month, before their third trimester and receive regular visits (frequency varies by developmental stage: weekly to monthly) from nurses during pregnancy through the child’s second birthday. The NFP National Service Office collects data on all enrolled families both at intake and at defined time points regarding visits, screenings, and outcomes (including maternal and child health). We have previously worked with these data and are well-positioned to access and analyze them. Our estimated sample size for the primary aim is 3000 families per group. This will allow sufficient power to conduct regression analyses, controlling for family, program site, and community characteristics. This project will provide critical new knowledge about the importance of starting the home visiting relationship in person vs. throug...