Background: This study with five-year-olds is the first to examine whether low-quality interactions with parents elicit physiological stress in children beyond toddlerhood, as evident from elevated cortisol levels in their saliva. It was hypothesised that particularly children with low levels of ego-resiliency--a personality construct reflecting the capacity to cope with stress--would show cortisol increases during low-quality parent-child interactions. Method: In a sample of 101 five-year-old children (62 boys), parent-child interaction was observed at home during parent-child discourse that involved the recollection and discussion of emotional events that happened to the child in the past. Saliva samples to assess cortisol levels were collected before and 20 minutes after the parent-child discourse task. The children's teacher rated child ego-resiliency using a Dutch translation of the California Child Q-set (CCQ; Block & Block, 1980). Results: One of the two parent-child interaction factors that emerged from a principal component analysis, namely negative parent-child interactions, was significantly related to the children's cortisol reaction; more negative parent-child interactions elicited significantly stronger cortisol reactions. The other parent-child interaction factor that was found, i.e., effective guidance, was not significantly related to children's cortisol reaction. As predicted, children low on ego-resiliency showed increases in cortisol during negative interactions with their parents, whereas high ego-resilient children did not. Conclusions: The association between negative parent-child interactions and cortisol elevations in children may point to a likely mechanism through which negative parent-child interactions contribute to negative developmental outcomes as the repeated exposure to high levels of cortisol have earlier been found to negatively affect children's development and functioning in various areas.