Latex photometric immunoassay for CA19-9 was clinically evaluated. Reference interval of the assay was determined as 37U/ml based on the data of 350 healthy donors. CA19-9 values were higher in females than in males (p < 0.01) and distribution of CA19-9 values varied significantly depending on the Lewis blood types of the subjects; namely, subjects with Lea+b- gave highest CA19-9 values followed by those with Lea-b+ and then those with Lea-b- (p < 0.001). In the patients with various malignant diseases, positive rates were high (> 60%) in the patients with pancreatic, hepatic, and colon cancer. Meanwhile, false positive rates were around 25-50% in the patients with pancreatitis, cholelithiasis, benign hepatic disorders and benign lung diseases. When cut-off value was set at 100U/ml, false positive rates decreased drastically, while true positive rates remained slightly decreased. These findings indicate that, in this latex assay, CA19-9 values of 100U/ml or higher should be regarded as high risk factor of malignancy and values in the range of > or = 37U/ml and < 100U/ml might be considered as warning signal of various benign and malignant diseases.