OBJECTIVE: To provide an epidemiological description of the use of ultrasound in obstetrics in Denmark and to analyse whether screening per se reduces the mean number of obstetric ultrasound examinations for all pregnant women. DESIGN: Questionnaire study based on hospital records and patient interviews. SETTING: All Danish hospital departments with delivery services (n = 57). PATIENTS: 2268 women who gave birth to a child in a two week period February 1990. RESULTS: The main findings of the study are: 1) Women with an offer of screening had a significantly higher mean number of obstetric ultrasound examinations during pregnancy (1.8) than women without (1.3), no matter at which type of department they had antenatal care and their delivery; 2) 45% of the women without an offer of screening had an ultrasound examination on indication in the interval of 14-20 weeks of gestation--at which time an examination may include what is performed at a screening examination and 3) The study shows great regional variations in the use of ultrasound in obstetrics. CONCLUSION: The results question the presumption that screening per se will reduce the mean number of ultrasound examinations during pregnancy.