This study tested whether the associations between interparental conflict, children’s emotional reactivity, and school adjustment were moderated by children’s cortisol reactivity in a sample of young children (N = 243; mean age = 4.6 years at Wave 1; 56% female, 44% male) and their parents. Using a longitudinal, autoregressive design, observational assessments of children’s emotional reactivity at Wave 2 mediated the relationship between an observational measure of Wave 1 conflict between parents and teacher’s report of children’s school adjustment at Wave 3. However, children’s cortisol reactivity to parent conflict at Wave 1 moderated the first link, such that emotional reactivity operated as a mediator for children with heightened cortisol reactivity but not children with low cortisol reactivity. Moderation was expressed in a “for better” or “for worse” form hypothesized by biological sensitivity to context theory. Thus, children with high cortisol reactivity experienced greater emotional reactivity than their peers when faced with more destructive conflict but also lower emotional reactivity when exposed to more constructive interparental conflict. Results are discussed as to how they advance emotional security and biological sensitivity to context theories.