To evaluate the accuracy of ultrasound in early detection of chorionicity and amnionicity in twin pregnancies and evaluate the role of operator-experience, a prospective follow up of 87 consecutive twin pregnancies was performed. Chorionicity and amnionicity were determined by 17 different observers, divided in 2 groups on the basis of their experience, using a composite of US markers at the first US examination confirming a viable intrauterine twin pregnancy. Accuracy of prediction was determined for both groups by comparison with postnatal pathology. From the 87 pregnancies examined at 10.1 (minimum 5.5 weeks, maximum 26.0) weeks of gestation, pathology was available in 82 cases and a correct prediction of chorionicity was made in all but two cases, both being diagnosed as monochorionic while dichorionic by observers belonging the more experienced group. No monochorionic pregnancy was missed and amnionicity was correctly diagnosed in all cases, including two monoamniotic twins. We conclude that use of a composite of ultrasound markers for early detection of chorionicity and amnionicity is reliable in a clinical setting where physicians with a variable degree of experience perform ultrasound examination.