Families randomly assigned to 1 of 4 treatment groups that varied across 2 orthogonal dimensions were compared at 1, 3, and 9 months to examine the effects of a neonatal intervention. In half of the families mothers and fathers were the target of intervention, and in the remaining half of the families mother was the sole target of intervention. The intervention itself consisted of passive or active exposure to the Brazelton Neonatal Behavior Assessment; half of the parents in each target-of-intervention group (mother only, mother and father) actively elicited responses from their newborn under the direction of a facilitator (active exposure) and the other half of the parents listened to a detailed verbal description of their infant's performance on the exam. Assessments of mother-infant interaction in the dyad and mother-infant, father-infant, and husband-wife interaction in the triad revealed no effects of the experimental intervention across groups. Discussion emphasizes the need to be cautious about the effectiveness of minor interventions and the need to be critical about past reports of success.