This study investigated whether transitory and persistent poverty spells were associated with children's learning (literacy and numeracy scores) and executive function outcomes in Ghana. Children resided in the Greater Accra region (N = 2,154; 49% female; M[subscript age] = 5.2 years at wave-1) and were followed at four-time points over three years. Multivariate regression models were used to test whether exposure to transitory poverty (1-2 times) or persistent poverty (3-4 times) were differentially associated with child learning and executive function outcomes at wave-4 (M[subscript age] = 7.5 years) compared to children who did not experience poverty at any wave. Our results show that transitory poverty was significantly and negatively associated with children's literacy scores. Persistent poverty had the largest negative relations for both learning outcomes. Findings were mixed for executive function, with persistent poverty negatively related to cognitive flexibility, but not inhibitory control. Implications for global child development and learning are discussed. [This is the online version of an article published in "Applied Developmental Science."]