Purpose : To determine the MR findings of cervical lymphadenopathy which distinguish tuberculouslymphadenitis(TL) from metastatic lymphadenopathy(ML). Materials and Methods : We retrospectively analyzed the MRfindings of 14 patients with TL and 19 with ML. Thirty-nine lesions in 14 patients with TL and 80 lesions in 19with ML were classified into three categories: conglomerate lesion (TL, 7; ML, 8); discrete lesion withoutnecrosis (TL, 3; ML, 15); and discrete lesion with necrosis (TL, 29; ML, 57). Between TL and ML, we compared age,sex, bilaterality, multiplicity, associated pulmonary tuberculosis, infiltration of adjacent fat, muscle, skin andsubcutaneous tissue, margin, and signal intensities on T1- and T2-weighted images. In discrete lesion, we alsocompared the minimal to maximal dimension ratio, and in discrete necrotic lesion, the thickness and evenness ofrim enhancement, and the location and extent of necrosis. Results : There were statistically significantdifferences between TL and ML in age(34.7$pm$13.1 years; 56.1$pm$14.3 years), male to female ratio(3:11; 13:6),bilaterality(1/14; 10/19), multiplicity(5/14; 14/19), and associated active pulmonary tuberculosis(7/14; 1/19). OnT2-weighted images, TL showed predominantly homogeneous (30/39) or high (20/39) signal intensity, similar to thatof cerebrospinal fluid, but the signal inteusity of ML was predominantly heterogeneous(58/80), or lower than orsimilar to that of fat(41/80). Between TL and ML, however, there were no statistically significant differences ininfiltration of adjacent tissue, margin, signal intensity and its heterogeneity on T1-weighted image, minimal tomaximal dimension ratio, thickness and evenness of rim enhancement, or the location and extent of necrosis.Conclusion : In cervical lymphadenopathy, the existence of TL rather than ML is suggested by single or unilaterallesion in a young woman, by associated active pulmonary tuberculosis, or by homogeneous high signal intensity onT2-weighted MR images.