Early life factors can shape the development of hypothalamic pituitary adrenal (HPA) axis. Maternal prenatal stress might constitute such an early environmental factor. As little is known about the relation between maternal prenatal stress and cortisol reactivity in human offspring, we performed a longitudinal study including four assessments of infant cortisol reactivity to stressful events in a non-clinical population. General and pregnancy-related feelings of stress and anxiety, as well as circadian cortisol levels, were measured in 173 mothers in the last trimester of pregnancy. Infant cortisol reactivity was measured at 5 weeks to a bathing session, at 8 weeks to a vaccination, at 5 months to a stressful mother-infant interaction (still face procedure), and at 12 months to a maternal separation (strange situation procedure). Maternal prenatal fear of bearing a handicapped child was a consistent predictor of infant cortisol reactivity. Although the effects were mild, higher fear was significantly related to higher salivary cortisol reactivity to the bathing session and to decreased cortisol reactivity to vaccination and maternal separation. Thus, pregnancy-specific anxieties predict infant cortisol reactivity in the first year of life, but the direction of the effect depends on infant age and/or the nature of the stressor. While this specific anxiety was a better predictor than stress experience or maternal cortisol concentrations, the underlying mechanisms of these associations remain unclear. Future studies should try to incorporate multiple measures of HPA-axis reactivity during development when studying the long-term consequences of maternal prenatal stress.