To determine the value of scintigraphic perfusion studies in evaluating pancreas transplant patients, we reviewed 56 of these studies in 22 patients who had 27 transplants. Seventeen patients underwent two or more studies. The perfusion studies were performed with 20 mCi (740 MBq) of 99mTc-DTPA injected as a bolus followed by eight to 16 serial 2-sec images and a 500,000-count immediate static image. Images were evaluated for (1) the time and intensity of pancreatic peak radioactivity relative to the time and intensity of the iliac arterial peak; (2) relative pancreatic to iliac arterial intensity on the static image; and (3) size, homogeneity, and definition of the pancreas. Clinical diagnoses at the time of scintigraphy of normal function (n = 36), rejection (n = 13), pancreatitis (n = 6), or arterial thrombosis (n = 1) were based on insulin requirement, urine amylase, serum glucose, serum amylase, response to therapy, cultures, CT, MR, sonography, scintigraphy with 67Ga or 111In-WBCs, percutaneous drainage results, angiography, surgery, and pathologic examination of resected transplants. Three 99mTc-DTPA perfusion studies showed no pancreatic perfusion, four showed decreasing perfusion on serial studies, and five showed progressive loss of definition of the pancreas on serial studies. Of the three patients with no detectable perfusion, one had a normally functioning transplant, one had arterial thrombosis with transplant infarction, and one had severe rejection with minimal function. Decreasing perfusion was associated with rejection in three patients and pancreatitis in one. Decreasing definition was seen in four patients with rejection and one with pancreatitis. We conclude that perfusion scintigraphy is useful, primarily when performed serially, although nonspecific for evaluating pancreas transplants.