Telomerase activity is intensively studied as a prognostic marker in malignancies, and generally detected by telomeric repeat amplification protocol (TRAP) assay. Recently, Ohyashiki et al. demonstrated a semi-quantitative procedure for TRAP assay, which is more useful than the qualitative one to analyze the correlation between telomerase activity and clinicopathological features in carcinomas. Therefore, we used semi-quantitative TRAP assay to analyze the correlation between telomerase activity and clinicopathological factors in colorectal carcinomas. Twenty-nine cases of advanced colorectal carcinoma, obtained by surgery, were studied. A telomerase detection kit, TRAP-eze, was used for the PCR based TRAP assay, and a DNA sequencer for a semi-quantitative analysis was used for a analysis. Telomerase activity was positively detected in all samples, and its mean value was 2.076+/-1.634 units (range; 0.536-8.932 units). Based on Dukes' classification, the activity tended to be higher in Dukes' A (2.722 units; P = 0.0525) and C (2.430 units) than in Dukes' B (1.552 units). The activity was higher in the right colon than at other locations. (right vs. left, rectum = 3.167 vs. 1.216 units; P<0.05, 1.604 units; P<0.01). The activity was higher in poorly differentiated adenocarcinoma than in well differentiated adenocarcinoma (3.743 vs. 1.491 units; P<0.05). Semi-quantitative TRAP assay is useful to analyze the correlation between telomerase activity and clinicopathological factors in colorectal carcinomas.