PURPOSE: To evaluate the MR imaging of Mullerian duct anomalies (MDA) and analyze its diagnostic accuracy associated gynecologic diseases and effect on treatment plan. MATERIALS AND METHODS: Twelve patients with suspected MDA and three with incidentally-found MDA were included. Axial, sagittal and coronal images of T1-, FSET2- or T2-, proton density- and Gadolinium-enhanced T1-weighted images were obtained with 1.5T (n=13) and 0.5T (n=2) MR machines. On the basis of Buttram and Gibbons' classification, MR images were analyzed and classified as laparotomy (n=7), hysteroscopy (n=7), laparoscopy (n=3) or hystersalpingography (n=2). These were analyzed in terms of external contour of the uterus, intercornual distance, signal intensity of the uterus and septum, associated genitourinary diseases, and the influence of MR imaging on treatment plan. RESULTS: A total of 15 cases of MDA were accurately classified by MR imaging, as follows: didelphyses (n=6: 40%); ageneses (n=3: 20%); septate uteri (n=3: 20%); bicornuate uteri(n=2: 13%) and unicornuate uterus(n=1: 7%). In 14 cases (93%) findings corresponded with those of other examinations. The external contour of the uterus was banana-shaped in the didelphyc and unicornuate uterus, a fundal cleft in the bicornuate, and normal in the septate; intercornual distance was greater in didelphyc and bicornuate uteri. Except in three cases of agenesis, the signal intensity of the uterus was normal; in the septum, on T2WI, this was isosignal in the bicornuate uterus and low in the septate, as compared with that of the myometrium. In four of five cases, associated gynecologic diseases were found on MR imaging; in five cases (33%), there were associated urologic anomalies, and in six (40%), MR imaging influenced the treatment plan. CONCLUSION: MR imaging was valuable for the diagnosis of MDA and the determination of associated gynecologic diseases, and also influenced the treatment plan.