Twenty-nine infants undergoing phototherapy to treat hyperbilirubinemia were divided into two groups. One group of mothers (non-separated) were transferred with their infants to the paediatric ward and remained as the primary care provider. The other group of mothers (separated) remained on the maternity ward and went to the paediatric unit to care for their infants at feeding times. A third group of healthy mothers and infants served as controls. Mothers in the non-separated group were more satisfied with the care than mothers in the separated group. The duration of breast feeding in the separated group was significantly shorter than in the other groups. The urine cortisol levels in the separated mothers were the same at one week and three months after delivery. The mothers in the other two groups had significantly higher urine cortisol levels during the first week postpartum than three months postpartum. We conclude that routine phototherapy for jaundiced infants should not separate mothers and infants.