Background Previous studies concerning symptom-giving pelvic girdle relaxation in pregnancy have to our knowledge been retrospective. We wanted to 1) determine the incidence during pregnancy and the prevalence two, six, and twelve months post partum, 2) identify possible predisposing factors, and 3) determine the frequency and duration of sicklisting, prospectively. Material and methods. A cohort of 1600 consecutive pregnant women filled in a questionnaire. At the routine prenatal examinations they were asked about pelvic pain. Those who fulfilled the inclusion criteria were examined by a rheumatologist to confirm the diagnosis. The affected women were seen again two, six, and twelve months post partum. All participants were asked about sicklisting in pregnancy. Results. The incidence during pregnancy was 14%, the prevalence two, six, and twelve months post partum were 5%, 4%, and 2% respectively. Multivariate analysis indicates that the most important predisposing factor is pelvic pain in a previous pregnancy. Other factors were uncomfortable working conditions, lack of exercise, and previous low back pain and low abdominal pain. At least 37% of the women with symptom-giving pelvic girdle relaxation had been sicklisted in pregnancy due to pelvic pain, on average for twelve weeks. Conclusion, Symptom-giving pelvic girdle relaxation is a considerable problem both in pregnancy and post partum. The occupational risk can possibly be prevented. The syndrome has a great social impact because of the frequent sicklisting.