In certain domains, people represent some of an individual's properties (e.g., a tiger's ferocity), but not others (e.g., a tiger's being in the zoo), as stemming from the assumed "essence" of the individual's category. How do children identify which properties of an individual are essentialized and which are not? Here, we examine whether formal explanations--that is, explanations that appeal to category membership (e.g., "That's ferocious 'because it's a tiger'")--help children to identify which properties are essentialized. We investigated this question in two domains: animal kinds (Study 1) and social categories (specifically, gender; Studies 2 and 3). Across studies, we introduced children to novel behaviors and preferences of individuals using either a formal explanation or closely matched wording that did not express a formal explanation. To measure the extent to which children essentialized the novel properties, we assessed their inferences about the stability, innateness, and generalizability of these properties. In Study 1 (N = 104; 61 girls, 43 boys; predominantly White and multiracial children from high-income backgrounds), we found that formal explanations led 5- and 6-year-old children to view novel properties of individual animals as more stable across time. In Studies 2 and 3 (total N = 163; 84 girls, 79 boys; predominantly White, Asian, and multiracial children from high-income backgrounds), we found that formal explanations led 6-year-olds, but not 5-year-olds, to view novel properties of individual girls and boys as more stable across contexts. These studies highlight an important mechanism by which formal explanations guide conceptual development.