The functions of the electroencephalographic rhythms are uncertain. Correlational evidence has linked 10-12Hz alpha rhythms to memory formation in healthy people. Moreover, loss of 10Hz alpha correlates with memory problems in Alzheimer's disease. Going beyond mere correlation, brain stimulation or peptides that alter electroencephalographic rhythms can modulate behaviour and enhance memory in rats. This latter finding fits with evidence that electroencephalogram-frequency stimulation can enhance long-term potentiation, the neural basis of memory. I aimed to test if manipulations of alpha-frequency electroencephalographic activity enhance human memory. Flicker provides an experimental means of modulating the human electroencephalogram: 9-12Hz flicker can entrain alpha-like activity. In rats, the frequency-specificity of brain stimulation's behavioural effects excludes the possibility that they result from non-specific (i.e. non-electroencephalographic) mechanisms. I tested if flicker manipulations would show analogous frequency-specific effects in man. In view of the above correlational data, I predicted that flicker at 10Hz (a frequency near to the peak power of endogenous alpha) would enhance human memory, but adjacent frequencies (8.7 and 11.7 Hz) would not. The results confirmed this prediction. This suggests that 10Hz electroencephalographic alpha subserves memory formation in man: 10Hz flicker enhances memory in healthy people and may have therapeutic potential in memory disorders.