In the print summary of this Research paper by Cooper and colleagues (BMJ 2009;338:b974, 25 April 2009, doi:10.1136/bmj.b974) an editorial oversight led to the authors’ final amendments not being incorporated. The “Design” section should have included the size of the intervention group (n=220) and should have specified that the authors used videotaped play observations to assess the quality of mother-child interactions at 6 and 12 months and infant attachment at 18 months post partum. The “Participants and setting” section should have specified that, of a consecutive series of 452 pregnant women, three refused to participate and 449 were recruited.