OBJECTIVE: To determine the usefulness of brain-derived neurotrophic factor (BDNF) as a diagnostic biomarker for colorectal cancer (CRC). MATERIALS AND METHODS: ELISA immunoassay was used to examine BDNF concentrations in the sera of two different retrospective cohorts consisting of CRC patients and age/gender matched controls. Cohort 1 consisted of 99 controls and 97 CRC patients, whereas cohort 2 consisted of 47 controls and 91 CRC patients. RESULTS: In cohort 1, the median concentration of BDNF was significantly (p< 0.0001) lower in CRC patient samples (18.8 ng/mL, range 4.0-56.5 ng/mL) than control samples (23.4 ng/mL, range 3.0-43.1 ng/mL). This finding was validated in an independent patient cohort (CRC patients: 23.0 ng/mL, range 6.0-45.9 ng/mL; control patients: 32.3 ng/mL, range 14.2-62.4 ng/mL). BDNF concentrations did not differ significantly between Dukes' staging in the patient cohort, however patients with Stages A, B, C and D (p< 0.01 for each stage) tumours had significantly reduced BDNF levels compared to healthy controls. Receiver operating characteristic analysis was performed to determine the ability of BDNF to discriminate between healthy controls and those with CRC. At 95% specificity, BDNF concentrations distinguished CRC patients with 25% and 18% sensitivity, respectively, in cohorts 1 and 2 (cohort 1: AUC=0.79, 95% CI 0.70-0.87; cohort 2: AUC =0.69, 95% CI 0.61-0.76). CONCLUSION: The serum levels of BDNF were significantly lower in colorectal cancer patients when compared to a control population, and this did not differ between different Dukes' stages.