To develop robust prediction models for infant obesity risk, we need data spanning multiple levels of influence, including child clinical health outcomes (eg, height and weight), information about maternal pregnancy history, detailed sociodemographic information of parents and community-level factors. Few data sources contain all of this information. This manuscript describes the creation of the Obesity Prevention in Early Life (OPEL) database, a longitudinal, population-based database that links clinical data with birth certificates and geocoded area-level indicators for 19 437 children born in Marion County, Indiana between 2004 and 2019. This brief describes the methodology of linking administrative data, the establishment of the OPEL database, and the clinical and public health implications facilitated by these data. The OPEL database provides a strong basis for further longitudinal child health outcomes studies and supports the continued development of intergenerational linked clinical-public health databases.