Mathematics performance in early schooling is linked to both later success in adulthood and, indirectly, to future health outcomes. Compared to reading, we know relatively little about the early predictors of later mathematical success and how to help children who are falling behind. Research has shown that children's working memory capacity (WMC) is one of the best cognitive predictors of mathematical achievement. Defined as the amount of information that can be processed and remembered simultaneously, WMC typically explains 25% of variance in mathematics performance. In a previous study, we showed that this correlation is particularly high at the start of formal schooling and surpasses that of prior performance in mathematics (Lee & Bull, 2016). Surprisingly, despite robust individual differences in WMC at each grade, the rate at which WMC increased did not differ across individuals. This is an important finding and put the spotlight on the origins of individual differences in WMC. The proposed study will examine whether some aspects of socioeconomic status (SES) are more closely associated with differences in WMC. It will also examine whether this relation is mediated by factors associated with SES (e.g., parenting practices, the kind of games to which children are exposed). Data will be collected from several sites worldwide because subjective perception of SES and its affordances will likely vary with different socio-cultural practices. Data collection just commenced in Hong Kong and will soon commence in Thailand. At each site, we will sample ~211 5-year-old children (stratified into low, medium, and high SES groups) and their main care-givers. Children will be administered a short battery of three working memory tasks. Their care-givers will be asked to complete a survey on SES parenting stress, parenting styles, and home affordances. [EdUHK_logo]