&NA;Serial estimations of serum Tennessee antigen have been performed at regular three‐month intervals on 35 patients with colorectal cancer who had undergone resection of all macroscopically obvious tumor but who were considered to be at high risk of developing subsequent metastases. The results were interpreted by a panel of surgeons in order to assess the clinical relevance of using serum Tennessee antigen for monitoring of patients. The serial estimation of serum Tennessee antigen was found to be very variable, difficult to interpret, and clinically unreliable as an accurate marker for the development of recurrent cancer in this group of patients. There are unacceptably high false‐positive and false‐negative diagnostic rates for serum Tennessee antigen estimations in comparison with serial estimations of carcinoembryonic antigen.