To understand the determinants of frightening/frightened and other atypical maternal behavior, the authors studied a sample of 41 inner-city mothers of very young children (ages 8-50 months), the mothers of whom had lifetime histories of interpersonal violent trauma (i.e., physical or sexual abuse, and domestic violence) and related posttraumatic stress. The authors measured (1) maternal salivary cortisol levels before and 30 minutes after a videotaped play paradigm with their children, involving two separations and reunions; and (2) cortisol reactivity 30 minutes after separation stress. Data were analyzed using Pearson bivariate correlations, ANOVA, and multiple linear regressions. Salivary cortisol "baseline" values were significantly negatively correlated with childhood interpersonal violent trauma severity (i.e., trauma severity prior to age 16). However, cortisol reactivity was not significantly correlated with interpersonal violent trauma severity at this level of analysis. Although baseline salivary cortisol values were not significantly correlated with current overall psychiatric or depressive symptoms, they were negatively correlated with severity of current posttraumatic stress symptoms (PTSS) and with dissociative symptoms. Neither dimensions of negativity nor distortion of maternal attributions showed any significant association with prestress or poststress salivary cortisol levels. Salivary cortisol baseline was negatively correlated with atypical maternal behavior via measurement of the level of disrupted communication, at a trend-level of significance. Violent trauma-associated dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis may be a marker for increased risk for intergenerational transmission via parenting behavior with young children. Low salivary cortisol prior to separation stress and blunted cortisol reactivity to separation may also be markers for posttraumatic stress.