We analyzed snoring noise from 10 nonapneic heavy snorers and nine patients with obstructive sleep apnea (OSA). Sound was recorded simultaneously through two microphones, one attached to the manubrium sterni and one suspended in the air 15 cm from the patient's head. Signals were stored on magnetic tape, digitized, and displayed in the time and frequency domains. Most of the power of snoring noise was below 2,000 Hz, and the peak power was usually below 500 Hz. When snores were generated during nose-only breathing (nasal snores), the sound spectrum was made up of a series of discrete, sharp peaks, with a fundamental note and harmonics similar to the spectrum of voiced sounds. When snores were generated during breathing through nose and mouth (oronasal snores), the spectra showed a mixture of sharp peaks and broad-band white noise. Patients with apnea showed a sequence of snores with spectral characteristics that varied markedly through an apnea-respiration cycle. The first postapneic snore consisted mainly of broad-band white noise with relatively more power at higher frequencies, so that the ratio of power above 800 Hz to power below 800 Hz could be used to separate snorers from patients with OSA. Other breaths in the cycle resembled oronasal or nasal snores in nonapneic subjects. Characteristics of the noise give information about the possible mechanism of sound generation and thus about the behavior of the pharynx during snoring. Quality of snoring sound may help to separate patients with obstructive apnea from those with simple snoring.