Sleep fragmentation is considered to have a central role in the pathophysiology of sleep-disordered breathing. However, the evaluation of sleep fragmentation can be very troublesome work. We sought to evaluate sleep fragmentation in 10 patients with obstructive sleep apnea syndrome by utilizing fast-fourier transforms to render their sleep encephalograms as spectrograms (horizontal axis: time, vertical axis: frequency, color: power). Detailed 5-minute spectrograms revealed relatively uniform patterns of change in EEG spectra during and after apnea episodes. In non-REM sleep, delta activity during apnea was followed by alpha and beta activity during resumed breathing. In REM sleep, flat and low-power spectral patterns during apnea were followed by relatively weak alpha and beta activity during resumed breathing. In compressed, 1-hour views of the spectrograms, repeated apnea events appeared as vertical striped patterns of spectral change. These patterns were thought to be representative of sleep fragmentation. The spectrographic display of EEGs may be a very simple and easy method for the evaluation of sleep-fragmentation in sleep disordered breathing.