Lymph node involvement in colorectal cancer, one of the most important prognostic factors, can be sometimes underestimated. In this study the authors report the results of two different techniques of specimen preparation and examination. In 240 patients (Group I), histologic examination was performed using a conventional procedure. In Group II (60 cases) the resected bowel and its mesentery were separately stretched, pinned on to a cork board, and fixed. The mesentery was divided according to node location (intermediate and principal) and evaluated by sight and palpation to identify lymph nodes. The bowel segment was divided from 5 cm proximally to 5 cm distally to the tumor every 10 mm in serial 3 mm slices. Three and 10 mm slices were then carefully examined by sight and palpation. Isolated lymph nodes embedded in groups (10-12 per paraffin block) were stained and investigated for neoplastic involvement. The specimen examination procedure used in Group II resulted in identification of a higher number of lymph nodes (mean = 41.1) and nodal metastases (mean = 10) compared to the standard technique used in Group I (mean = 11.3 and 2.4, respectively--P < .05). The percentage of N+ cases also was increased in Group II (48.3%) when compared to that in Group I (30.4%; P < .05). The new technique is simple, inexpensive, and efficacious for the detection of lymphatic metastases in colorectal cancer.