Trials investigating colorectal cancer (CRC) chemoprophylaxis with cyclooxygenase-2 (COX-2) inhibitors have been discontinued because of adverse cardiovascular effects. Nevertheless, identification of patients where beneficial, chemo-prophylactic effects of COX-2 inhibitors outweigh side-effects may be possible; this study aimed to investigate whether such patient groups might exist. The COX-2 status of viable epithelial and inflammatory cells in freshly disaggregated CRC and paired normal colonic samples was assessed by three-colour flow cytometry. 21/31 (67.7%) CRCs expressed COX-2, with inflammatory cells positive in 19/31 (61.3%), epithelial cells in 12/31 (38.7%), and both positive in 10/31 (32.3%). 25/30 (83.33%) normal samples expressed COX-2, with epithelial cells positive in 18/30 (60%), inflammatory cells in 15/30 (50%) and both positive in 10/30 (33.3%). Strength of expression by CRC and normal was similar. More advanced cancers had higher expression rates (COX-2 in 12/13 (92.3%) with nodal disease vs 9/17 (52.9%) node-negative; P = 0.04). Investigation of ex-vivo CRC cells by flow cytometry demonstrated COX-2 expression rates comparable to that previously reported. However, expression by paired live normal colon was significantly greater, suggesting that COX-2 may be expressed at higher rates in normal colonic cells in patients with CRC. Patients identified at resection as expressing COX-2 in normal colon may benefit from Coxib chemo-prophylaxis, thus potentially offering a refined approach to that adopted in the VICTOR trial.