Using a nationally representative dataset (Early Childhood Longitudinal Study-Birth Cohort) and bioecological-cumulative disadvantage framework, the present study investigated the relations between salient child and family risk experiences and American Indian and Alaska Native (AIAN) children's kindergarten academic and behavioral outcomes. Using hierarchical linear regression, individual risk models, cumulative risk models, and interaction models examining combinations of two types of individual risks were examined. Individual risk models including seven risks (i.e., poverty, preterm/low birth weight, low maternal education, single motherhood, inadequate prenatal care, teen motherhood, and severe maternal depression) revealed that poverty exposure at any point prior to kindergarten was associated with AIAN children's kindergarten academic skills (i.e., reading and math). Cumulative risk models suggest that children exposed to two or more risks displayed lower reading and math skills than those exposed to no or one risk. The interaction models revealed significant risk by risk interaction effects for kindergarten math and all behavior outcomes examined in this study. Specifically, children who experienced poverty and had mothers with risk characteristics (i.e., without a high school diploma or who gave birth as teenagers) demonstrated poorer behavioral outcomes (i.e., lower social competence and approaches to learning as well as higher externalizing behaviors) than children experienced poverty but with mothers having no such risk characteristics. Interestingly, findings also revealed that children living in poverty presented better kindergarten outcomes (i.e., scored higher on math and approaches to learning and lower on externalizing behaviors) when they were living with single mothers than with married/cohabitating parents. Given the salience of specific combinations of poverty and maternal characteristics for AIAN children, implications for two-generation programming is discussed along with the potential value of extended family networks. Further study implication was discussed. (author abstract)