Postpartum perineal repair performed by midwives: A randomised trial comparing two suture techniques leaving the skin unsutured. Objective To compare a continuous suture technique to interrupted stitches using inverted knots for postpartum perineal repair of second-degree lacerations and episiotomies. Design A double blind randomised controlled trial. Setting A Danish university hospital with more than 4800 deliveries annually. Population 400 healthy primiparous women with a vaginal delivery at term. Method Randomisation was computer-controlled. Structured interviews and systematic assessment of perineal healing was performed by research midwives blinded to treatment allocation at 24 to 48 hours, ten days and six months postpartum. Pain was evaluated using a Visual Analogue Scale and the McGill Pain Questionnaire. Wound healing was evaluated using the REEDA scale and by assessment of gaping wounds >0.5 cm. Analysis complied with the intention-to-treat principle. Main Outcome measures The primary outcome was perineal pain ten days after delivery. Secondary outcomes were wound healing, patient satisfaction, dyspareunia, need for resuturing, time elapsed during repair and amount of suture material used. Results 400 women were randomised; five withdrew consent leaving 395 for follow-up. The follow-up rate was 98% for all assessments after delivery. No difference was seen in perineal pain ten days after delivery. No difference was seen in wound healing, patient satisfaction, dyspareunia or need for resuturing. The continuous suture technique was significantly faster (15 min. vs. 17 min, p=0.03) and less suture material was used (1 vs. 2 packets, p<0.01). Conclusion Interrupted, inverted stitches for perineal repair leaving the skin unsutured appears to be equivalent to the continuous suture technique in relation to perineal pain, wound healing, patient satisfaction, dyspareunia and need for resuturing. The continuous technique, however, is faster and requires less suture material thus leaving it the more cost-effective of the two techniques evaluated.