OBJECTIVE: An earlier publication showed that acupuncture and stabilising exercises as an adjunct to standard treatment was effective for pelvic girdle pain during pregnancy, but the post-pregnancy effects of these treatment modalities are unknown. The aim of this follow-up study was to describe regression of pelvic girdle pain after delivery in these women. DESIGN: A randomised, single blind, controlled trial. SETTING: East Hospital and 27 maternity care centres in Göteborg, Sweden. POPULATION: Some 386 pregnant women with pelvic girdle pain. METHODS: Participants were randomly assigned to standard treatment plus acupuncture (n=125), standard treatment plus specific stabilising exercises (n=131) or to standard treatment alone (n=130). MAIN OUTCOME MEASURES: Primary outcome measures: pain intensity (Visual Analogue Scale). Secondary outcome measure: assessment of the severity of pelvic girdle pain by an independent examiner 12 weeks after delivery. RESULTS: Approximately three-quarters of all the women were free of pain 3 weeks after delivery. There were no differences in recovery between the 3 treatment groups. According to the detailed physical examination, pelvic girdle pain had resolved in 99% of the women 12 weeks after delivery. CONCLUSIONS: This study shows that irrespective of treatment modality, regression of pelvic girdle pain occurs in the great majority of women within 12 weeks after delivery.