Background: Having health rankings at the ZIP code level could allow more effective engagement of community stakeholders and would facilitate further research on the effect of small-area variation in health factors on health outcomes. Our project develops a novel county-level measure of population health using data that are available at the ZIP code level for the purposes of extending county-level health index insight to sub-county level geographies. Methods: Using the same conceptual framework as the County Health Rankings (CHR), ZIP and county level CHR domain scores were derived for Missouri from hospital discharge data, state and national health survey data, and other public and commercial datasets. Two variants of the Health Outcomes domain score analog were constructed: one based on the derived, shrunken quality of life (QOL) and mortality subdomain scores combined using CHR composite weights; another augmented version also included Health Factors scores in the Health Outcomes domain derivation. We then assessed the agreement between aggregated county-level results and CHR indices using both of these models. Results: Correlations between CHR overall domain scores and derived analogs using standard CHR weights for 115 Missouri counties were .87 (p<.001) for the Health Factors domain and .67 (p<.001) for Health Outcomes domain. Using the augmented model produced a correlation of .83 (p<.001) for the Health Outcomes domain. Discussion: These results suggest that reasonable approximations of county-level health indices can be derived with data that are widely available at sub-county levels and be reliably extended to the zip code level.