Importance: Virtual home-based interventions may bolster protective factors, such as family health and resilience, during stressors such as the COVID-19 pandemic; however, their effectiveness is unknown. Objective: To examine the effectiveness of a virtual health coaching intervention on family health and resilience during the pandemic. Design, Setting, and Participants: In this parallel-group, single-site randomized clinical trial, 123 parents and their 2- to 8-year-old children were enrolled at a pediatric clinic or community partner site in Tennessee from March 10 to August 11, 2021. Follow-up surveys were completed between June 29 and November 11, 2021. Interventions: All participants received 11 weekly cooking videos and associated home-delivered groceries. The intervention group also received 12 weekly, 30-minute virtual health coach sessions. Main Outcomes and Measures: The primary outcome was the validated 6-item (range, 6-30) Family Healthy Lifestyle Subscale (FHLS) scores. The secondary outcome was the validated 6-item (range, 0-6) Family Resilience and Connection Index (FRCI) scores. Outcomes were determined a priori and evaluated at baseline and 12-week follow-up. A priori independent t tests and multivariable tobit regression models assessed intervention effects, and post hoc, secondary interaction models assessed whether effects differed over baseline outcomes. Results: Among the 123 enrolled families, 110 (89%) were included in the primary analyses (parent mean [SD] age, 35.1 [8.2] years; 104 [95%] female; 55 [50%] non-Hispanic Black; child mean [SD] age, 5.2 [1.7] years; 62 [56%] male). Intervention-control group mean differences were nonsignificant for follow-up FHLS scores (0.7; 95% CI, -0.6 to 2.0; P = .17) and FRCI scores (0.1; 95% CI, -0.5 to 0.6; P = .74). Tobit regression model intervention effects were nonsignificant for FHLS scores (0.9; 95% CI, -0.3 to 2.2; P = .15) and FRCI scores (0.4; 95% CI, -0.2 to 1.1; P = .17). Post hoc, secondary models found no significant interaction for FHLS scores (1.3 increase per 5-point decrease; 95% CI, -0.2 to 2.7; P = .09), with significant intervention associations for baseline scores of 6 to 23. The interaction was significant for FRCI scores (0.4 increase per 1-point decrease; 95% CI, 0.01 to 0.8; P = .047), with significant intervention associations for baseline scores of 0 to 3. Conclusions and Relevance: In this randomized clinical trial of families with young children, weekly virtual health coaching did not detectably improve family health and resilience. Post hoc, secondary results provided preliminary evidence of potential effectiveness among families with low baseline scores. Trial Registration: ClinicalTrials.gov Identifier: NCT05328193.