In this study the ability of the micro-leucocyte adherence inhibition assay to detect pancreatic cancer in a population of 212 individuals was evaluated. Buffy coat leucocytes of 60 pancreatic carcinoma patients, 33 benign controls including 17 patients with symptomatic pancreatitis, 89 malignant controls, and 30 normal healthy volunteers were tested one or more times against extracts of pancreatic and control colorectal and gastric carcinomas. Tumor-specific immune responses to the pancreatic tumor extracts were detected in 11/12 patients with localized disease and 10112 patients with metastatic disease to sites other than the liver. However, once the primary pancreatic tumor metastasized to the regional lymph nodes or the liver, the diagnostic accuracy of the assay was reduced to 58% (21/36). The false-positive rate was less than 5% (7/152) with all control patients. No patients with symptomatic pancreatitis reacted to the pancreatic organ-specific neoantigen(s). Preliminary studies were performed to determine if a modified LA1 assay could be performed with patient's serum instead of patient's buffy coat leucocytes. Fifty-four serum samples from 13 patients with pancreatic cancer, six patients with gastric cancer, eight malignant and benign controls, and 16 normals were used to "arm" normal peripheral blood leucocytes and tested against extracts of pancreatic and gastric cancer. Serum from pancreatic cancer patients could "arm" normal leucocytes to recognize the pancreatic tumor extract in 13/21 tests, whereas control sera could "arm" normal leucocytes to recognize the pancreatic tumor extract in 1/33 tests. In contrast, only serum from gastric cancer patients could "arm" normal leucocytes to recognize the gastric tumor extract in 5/9 tests. Correlations between the serum-arming LA1 test and standard LA1 test indicated concordance in 24/31 tests. These results suggest that specific antitumor immunity can be detected with patients' serum and the activity of patients' buffy coat leucocytes may be due to a serum "arming" factor. Both the standard LA1 test and modified serum LA1 tests are worthy of further study in the differential diagnosis of pancreatic cancer.