ABSTRACT: BACKGROUND: Recent clinical studies have indicated the beneficial impact of dietary flavonoid intake on human cognitive performance. Although the mechanisms that mediate such improvements are currently unclear, animal and human trial data suggest that changes in neurotrophin expression, and related signalling apparatus, may be involved. OBJECTIVE: To investigate the link between changes in serum brain-derived neurotrophic factor (BDNF) and changes in human cognitive performance following flavonoid intake. METHODS: The relationship between serum levels of BDNF and age, gender, BMI, waist circumference, blood pressure and cognition at baseline, and following flavonoid intake, was investigated in two distinct randomised, controlled clinical trials. Trial 1 was conducted in men and women (aged 26-70?y; consuming an average of 3 portions of fruit and vegetables per day) and delivered high-flavonoid (>15?mg/100?g) or low-flavonoid (<5?mg/100?g) fruit and vegetables and increased intake by 2 portions every 6 weeks. The control arm was habitual diet over the same time frame. Trial 2 was conducted in an older males and female cohort (aged 62-75?y) intervening with a high-flavanol cocoa drink (494?mg total flavanols) and a low-flavanol cocoa drink (23?mg total flavanols) for 12 weeks. RESULTS: Serum BDNF levels increased linearly to the age of 65, after which BDNF levels were found to decrease markedly. All other physiological and anthropometric measurements proved to not be significantly associated with serum BDNF levels (p?>?0.05), although higher levels in males compared to females almost achieved significance (p?=?0.056). At baseline, higher serum BDNF levels were associated with significantly better global cognition scores, relative to individuals with lower serum levels. In addition, following intervention for 18 weeks, high-flavonoid, but not low-flavonoid, fruit and vegetable intake induced significant improvements in cognitive performance and increases in serum BDNF levels (p?=?<0.001). Flavanol intervention for 12 weeks also resulted in significant increases in serum BDNF (p?=?<0.001), and such increases were correlated with improvements in global cognitive performance. CONCLUSION: Increases in global cognition induced by high flavonoid fruit and vegetables, and cocoa flavanols were paralleled by concurrent changes in serum BDNF levels, suggesting a role for BDNF in flavonoid-induced cognitive improvements. Furthermore, we provide further data suggesting that serum BDNF levels may represent a biomarker of cognitive function.