Animal studies demonstrate that maternal touch and contact regulate infant stress, and handling during periods of maternal deprivation attenuates the stress response. To measure the effects of touch on infant stress reactivity during simulated maternal deprivation, 53 dyads were tested in two paradigms: still-face (SF) and still-face with maternal touch (SF+T). Maternal and infant cortisol levels were sampled at baseline, reactivity, and recovery and mother's and infant's cardiac vagal tone were measured during the free play, still-face, and reunion episodes of the procedure. Cortisol reactivity was higher among infants in the SF condition and while cortisol decreased at recovery for infants in the SF+T, it further increased for those in the SF. Vagal tone showed a greater suppression when SF was not accompanied by maternal touch. Touch synchrony during free play was associated with higher infant vagal tone, whereas touch myssynchrony--maternal tactile stimulation while the infant gaze averts--correlated with higher maternal and infant cortisol. In humans, as in mammals, the provision of touch during moments of maternal unavailability reduces infants' physiological reactivity to stress.