PURPOSE: To determine the utility of dynamic magnetic resonance (MR) imaging in evaluation of head and neck lesions. MATERIALS AND METHODS: Dynamic MR imaging was performed in 79 head and neck lesions. Five types of time versus signal intensity curves were obtained: type A, curve peaks 0-30 seconds after administration of contrast material; type B, curve peaks at 30-60 seconds; type C, curve peaks at 60-210 seconds; type D, gradual upward slope; and type E, flat. RESULTS: Use of dynamic imaging did not improve the sensitivity of nondynamic imaging for malignancy. Although malignancies had type B, C, or D curves, 39 (68%) of 57 benign lesions had the same patterns. Lesions with type A or E curves and lesions with type D curves and with marked hyperintensity on T2-weighted images were all benign. CONCLUSION: Although use of dynamic MR imaging contributes little in predicting malignancy in head and neck lesions, it can limit the differential diagnosis and has potential in predicting vascularity and recurrence.