Summary. A randomized controlled trial of two environments for delivery was conducted at Queen Charlotte's Maternity Hospital. A total of 253 parous women expecting to have a labour ward delivery were invited to participate in the trial but only 148 agreed. These women were randomly allocated to be delivered either with standard labour ward management (n= 72) or in the birthroom—a small bedroom decorated in a homely manner, without facilities for epidural analgesia or electronic fetal monitoring (n= 76). Eleven women in the birthroom group and 10 in the labour ward group withdrew from the trial before labour and four were transferred from the birthroom to the labour ward when in labour. A questionnaire sent in the postnatal period to the women who completed the trial was returned by 80%. In the birthroom group there was significantly (i) decreased admission-to-delivery interval (ii) less analgesia (iii) more freedom of movement (iv) less suturing (v) increased rooming-in. No difference was found in the assessment of difficulty of labour nor in the method of subsequent infant feeding.