Attachment security is theorized to shape stress reactivity, but extant work has failed to find consistent links between attachment security to mothers and infant cortisol reactivity. We examined family configurations of infant-mother and infant-father attachment security in relation to infant cortisol reactivity. One-year old infants (N = 180) participated in the Strange Situation with mothers and fathers in two counterbalanced lab visits, one month apart (12 and 13 months). Infants with secure attachments only to their fathers and not their mothers had higher cortisol levels than infants with a secure attachment to mother and also exhibited a blunted cortisol response (high at baseline and then a decrease after stress). Results suggest that a secure attachment to father may not be enough to reduce infant stress reactivity when the infant-mother attachment is insecure, and future research is needed to uncover the family dynamics that underlie different family configurations of attachment security.