To determine if glucocorticoid receptor levels could be used to predict response of patients to glucocorticoid therapy we have initiated a series of studies on patients and normal volunteers. Our major conclusions to date are: (1) All lymphomas appear to have measurable numbers of glucocorticoid receptors pretreatment. (2) Lymphoma cells have significantly more receptors than lymphocytes from non-malignant lymph nodes. (3) Among lymphomas there is a wide range in receptor number which is not predicted by histology. (4) Tumor receptor levels obtained simultaneously from different tissues in the same patient often vary. (5) In vivo administration of glucocorticoid rapidly results in a fall in receptor levels in lymphoid tissue which persists for as long as 17 days; thus patients should not have received glucocorticoid therapy for at least three weeks before receptors are assayed. (6) Tumor receptor levels at relapse are usually lower than at diagnosis in patients who have previously received glucocorticoid therapy and responded. (7) Tumor glucocorticoid receptor levels are significantly higher in patients who respond to a short course of glucocorticoid than they are in patients who fail to respond. Our studies suggest that glucocorticoid receptor determinations may allow us to select pretreatment those lymphoma patients who should receive glucocorticoid as part of their therapy.