PURPOSE: The purpose of this study was to evaluate the usefulness of contrast enhancement in assessing thedepth of stromal invasion in patients with uterine cervical tumors by comparing dynamic and late contrast-enhancedT1-weighted MR imaging with T2-weighted MR imaging. MATERIALS AND METHODS: Of 58 surgically proven uterinecervical cancer patients, 31 in whom tumors were seen on MRI were included in this study. Using a 1.5 T magnet,T2-weighted contrast-enhanced dynamic, and T1-weighted MR imaging were performed. In each MR imaging sequence,tumor visualization, margin delineation, enhancement pattern and depth of stromal invasion were evaluated onsagittal images and were correlated with pathological findings of resected uterus, focusing on the depth ofstromal invasion. RESULTS: Surgical FIGO stages were IB1 in 20 patients, IB2 in three, IIA in six, and IIB intwo. A tumor was detected in 29 (94%) patients on T2-weighted images, in 26 (84%) on dynamic contrast enhancedimages, and in 28 (90%) on contrast enhanced T1-weighted images. The tumor demonstrated a clear margin in 22 (71%)patients on late contrast-enhanced T1-weighted images, in 21 (68%) on dynamic images, and in 13 (42%) onT2-weighted images. Correlated with pathologic findings, the depth of stromal invasion was overestimated onT2-weighted images in eight (26%) patients, and on dynamic and late contrast-enhanced T1-weighted images in three(10%) and three (10%) respectively. CONCLUSION: Tumor margins are clearer and the extent of tumors may be moreaccurately evaluated on dynamic MR and late contrast-enhanced T1-weighted imaging. These sequences thas seem to beuseful, and superior to T2-weighted imaging, for assessing the depth of stromal invasion in patients with cervicalcarcinoma.