The results of combined CA-19-9 CEA assay were measured in 216 cases of colorectal cancer. In 28 preoperative patients, the positive rate 16.67% in Dukes' A group, 25% in B, 55% in C and 40% in D. It was proved less valuable in early diagnosis. The positive CA-19-9 alone in 66 with relapsed or metastases out of 182 undergoing radical resection was 63.63%, CEA alone 62.12%, and combined assay 86.36%. The false positive rate of CA-19-9 and CEA was 6.03% and combined assay 11.21%. In 27 palliative resections CA-19-9 in 40.74% cases, CEA in 44.44%, and combined assay in 59.2% was positive. In cases of nonresectable tumors, the positive rate was 66.7%, 66.7% and 83.33%, respectively. There was no definite correlation between the value of CA-19-9 and CEA. The data showed significantly higher sensitivity in combined assay than in either CA-19-9 or CEA alone. Combined assay with the sensitivity of 86.36% and the specificity of 88.79%, was more useful in finding of postoperative. recurrences or metastases. We suggest that this method should be used routinely in monitoring postoperative patients with colorectal cancer.