BACKGROUND: Stresses imposed by parenthood can provoke or intensify relationship problems between parents. These problems, which are often associated with postnatal depression, can have serious consequences for family well-being but are often not revealed to primary health care personnel. AIM: To evaluate a means of extending the primary health care team's ability to identify and respond to relationship problems of mothers and their partners in the postnatal period. DESIGN OF STUDY: Cluster randomised controlled trial. SETTING: Specially trained health visitors in nine 'intervention' clinics--each matched with a 'control' clinic' in an outer London borough. METHOD: Health visitors in intervention clinics invited mothers attending for the six-to-eight-week developmental check to complete a screening scale for relationship problems, and offered help (supportive listening, advice, or referral) if needed. When visiting the clinic for the 12-week immunizations, mothers from all clinics were asked to complete a follow-up self-report questionnaire. After the completion of the trial, 25 women who had attended the intervention clinics and had been offered support with a relationship problem were interviewed to elicit their views on the acceptability and value of the intervention. All 25 of the health visitors engaged in the intervention were asked to complete a questionnaire on their experience. RESULTS: Screening led to striking differences between intervention and control clinics in the percentage of women identified at the six-to-eight-week check as potentially in need of help with a relationship problem (21% versus 5%, P = 0.007) and in the percentage actually offered help (18% versus 3%, P = 0.014). About one-half of the mothers so identified were also identified as having postnatal depression. At the 12-week visit for immunizations, the intervention group was twice as likely (P = 0.006) as the control group to report having discussed relationship problems with the health visitor and 75% more likely (P = 0.046) to report having received help with a problem. CONCLUSION: The intervention offers a useful way of extending the primary health care team's ability to respond to problems that often have serious consequences for family well-being.