The purpose of this study was to evaluate the relative validity of responses to three different questions about snoring as indicators for sleep apnea in a population referred to a sleep clinic. Secondary goals were to evaluate the meaning of a "don't know" response to these questions and to examine how the associations between snoring and sleep apnea are influenced by demographics. Results from 1,409 patients in a sleep clinic indicated that nearly all levels of estimated snoring frequency were associated with a greater likelihood of sleep apnea. In addition, a "don't know" response indicated a likelihood of sleep apnea. In the sample from this clinic, sensitivities approximating 90 percent were obtained in men, and specificities approximating 90 percent were obtained in women, but high diagnostic accuracy (high specificity in men; high sensitivity in women) could not be achieved with the three snoring questions used here. Generally, associations between snoring and sleep apnea were independent of age and sex. Single persons, persons living alone, and persons customarily sleeping alone of both sexes all showed associations between self-reported snoring and the presence of sleep apnea.