Management of patients with stage II colorectal carcinomas remains challenging as 20 - 30% of themwill develop recurrence. It is postulated that these patients may harbour nodal micrometastases which areimperceptible by routine histopathological evaluation. The aims of our study were to evaluate (1) the feasibility ofmultilevel sectioning method utilizing haematoxylin and eosin stain and immunohistochemistry technique withcytokeratin AE1/AE3, in detecting micrometastases in histologically-negative lymph nodes, and (2) correlationbetween nodal micrometastases with clinicopathological parameters. Sixty two stage I and II cases with atotal of 635 lymph nodes were reviewed. Five-level haematoxylin and eosin staining and one-level cytokeratinAE1/AE3 immunostaining were performed on all lymph nodes retrieved. The findings were correlated withclinicopathological parameters. Two (3.2%) lymph nodes in two patients (one in each) were found to harbourmicrometastases detected by both methods. With cytokeratin AE1/AE3, we successfully identified four (6.5%)patients with isolated tumour cells, but none through the multilevel sectioning method. Nodal micrometastasesdetected by both multilevel sectioning and immunohistochemistry methods were not associated with larger tumoursize, higher depth of invasion, poorer tumour grade, disease recurrence or distant metastasis. We conclude thatthere is no difference between the two methods in detecting nodal micrometastases. Therefore it is opined thatmultilevel sectioning is a feasible and yet inexpensive method that may be incorporated into routine practice todetect nodal micrometastases in centres with limited resources.