Prenatal stress has a well-established link to negative biobehavioral outcomes in young children, particularly for girls, but the specific timing during gestation of these associations remains unknown. In the current study, we examined differential effects of timing of prenatal stress on two infant biobehavioral outcomes [i.e., hypothalamic-pituitary-adrenal (HPA) axis reactivity and difficult temperament] that are early-life precursors to the development of psychopathology. We obtained the most granular assessment of prenatal stress to date involving weekly stress ratings from 396 pregnant women between 15 and 41 weeks gestation. At 6 months postpartum, infant salivary cortisol was collected (n = 173) before and after a stressful laboratory task and mothers reported on infant temperament (n = 244). Machine learning explored both between- and within-person regression effects of prenatal stress on the two infant biobehavioral outcomes. For HPA axis reactivity, we found a sensitive period during mid-gestation (weeks 20 and 29) for girls, but during late gestation (week 37) for boys. For difficult temperament, we found a between-persons effect of mean stress level as well as sensitive periods in mid (weeks 20, 21, 25) and late gestation (week 37) for girls, but across mid to late gestation (weeks 25, 27, 30, 34, 40) for boys. This study is the first to use a weekly assessment across gestation to demonstrate specific windows of sensitivity for infant biobehavioral precursors of child psychopathology. The findings highlight that biological sex critically influences specific timing of these prenatal stress associations with infant outcomes, thus informing our understanding of sex differences in early biobehavioral markers of psychopathology.