Objective: Our purpose was to evaluate the risk of malignancy in surgically removed ovarian cysts that were characterized before the operation as unilocular according to transvaginal ultrasonography. Study Design: This prospective analysis included 927 premenopausal women and 377 postmenopausal women operated on at 2 European university hospitals between January 1992 and December 1997. On the basis of ultrasonographic findings the cysts were characterized either as echo-free, without solid parts or papillary formations (group 1), or as having echogenic cyst content, with solid parts or papillary formations (group 2). Ultrasonographic and macroscopic appearances of the cysts were compared with histopathologic diagnosis. Results: In group 1, in premenopausal women 3 of 413 cysts (0.73%) proved to be borderline or malignant, and in postmenopausal women 4 of 247 cysts (1.6%) proved to be borderline or malignant. The figures for cysts in group 2 were 11 of 514 cysts (2.1%) and 13 of 130 cysts (10.0%), respectively. It was not possible to differentiate by transvaginal ultrasonography between benign, borderline, and malignant cysts when solid parts or papillary formations were visualized. Conclusions: This study confirmed that the risk of malignancy associated with unilocular echo-free cysts (group 1) was low. Serial ultrasonographic follow-up should therefore be the standard procedure with unilocular echo-free cysts <50 mm in diameter. In cysts with a mean diameter of >50 mm, papillary formations or solid parts may be missed by transvaginal ultrasonography. The risk for malignancy in cysts containing papillary formations or solid parts (group 2) was 3 to 6 times higher than that in unilocular echo-free cysts. (Am J Obstet Gynecol 2001;184:48-54.)