Brain-derived neurotrophic factor (BDNF) is one of the neurotrophins promoting cognitive function and contributing to neurogenesis and neuroprotection. Available evidence suggests that exercise influences serum BDNF concentrations, but that the effect is transient. The purpose of this study is to determine whether a period of aerobic training, followed by a period of detraining, can influence basal serum BDNF levels in humans. Sixteen young, sedentary subjects were assigned to an experimental group (n=9) and a control group (n=7). The experimental group performed an aerobic training program during 8 weeks, followed by 8 weeks of detraining, during which subjects returned to their previous, sedentary activity level. The control group remained physically inactive during 16 weeks. In both groups, performance on short-term (Digit Span test) and mid-term memory (Recall of Images) was assessed. Aerobic training significantly increased the VO(2) peak in the experimental group, and these values returned to baseline after 8 weeks of detraining. Basal serum BDNF was not influenced by 8 weeks of aerobic training and detraining did not seem to have an effect on basal peripheral BDNF concentrations. Both training and detraining did not clearly influence short-term memory performance on the Digit Span test and no differences were present between the experimental and control group on the mid-term memory test. Future studies should focus on patient groups and elderly to further investigate the effect of training and detraining on neurotrophic factors and cognitive function, and on the effects of training and detraining on the BDNF response to acute exercise.