Little is known about predictors of publicly funded early care and education (ECE) use among low-income children of immigrants. Without this knowledge, it is difficult to effectively increase participation in these public programs, which promote school readiness but are underused by children of immigrants. Using nationally representative data from the Early Childhood Longitudinal Study -Birth Cohort (ECLS-B), this study attempts to identify pertinent family, child, maternal ECE preference, broader contextual, and immigrant specific characteristics predictive of ECE use among 4-year old children in a sample of low-income children of immigrants (N ≈ 1,050). Specifically, we estimate multinomial logistic regression models predicting type of ECE (Head Start, public pre-k, subsidized ECE, unsubsidized ECE, parental care) from these characteristics. Findings suggest that even in a low-income sample, correlates of disadvantage such as low maternal education and prior receipt of public benefits are important predictors of public ECE use, as are maternal preferences for certain features of care and supply-side factors such as ECE availability. Immigrant-specific factors such as English proficiency, citizenship status, availability of non-English speaking caregivers, and generosity of state policies toward immigrants emerged as particularly salient for explaining the public ECE selection patterns of low-income immigrants. Results point to future research areas and potential policy solutions aimed at increasing public ECE use for children who may stand to benefit the most.