Bone marrow suppression is often the limiting factor in the use of radiation therapy. In order to determine if MR imaging can be used to quantify bone marrow changes, we performed a serial prospective study of patients with lung cancer (six cases) and lymphoma (six cases). Quantitative and qualitative assessments of T1-weighted sagittal images, 750/33 (TR/TE), obtained at 0.6 T before, during, and after radiotherapy showed increased signal intensity in the radiated portions of the spine. These changes appeared as early as 2 weeks after the beginning of radiation, continued to increase until a maximum value was attained, and then persisted during the follow-up period of 2 years. A significantly higher (p less than .04) ratio of pretreatment to maximum posttreatment signal intensity was seen in patients with lymphoma than in those with lung cancer, and pretreatment values in patients with lymphoma were significantly lower (p less than .01). The lower pretreatment values found in the patients with lymphoma may have been due to the smaller amount of yellow marrow in these patients, who were significantly younger (33 vs 62 years). The higher ratio of pre- and posttreatment signal intensity may have been related to the larger amount of hematopoietic marrow available to undergo fatty replacement. The persistence of elevated signal intensity for as long as 2 years after radiation suggests an endpoint in the process of marrow conversion, but not reversal in the form of regeneration of hematopoietic bone marrow. Quantitative MR evaluation of bone marrow may be of considerable value as a noninvasive means of monitoring the effects of radiotherapy.