Prospective epidemiologic data of the American Cancer Society disclosed that reported usual sleep durations among groups who complained of insomnia and sleeping pill use "often" overlapped with those of groups who had no complaints. Reports of insomnia were not consistently associated with increased mortality when several factors were controlled; however, men who reported usually sleeping less than four hours were 2.80 times as likely to have died within six years as men who reported 7.0 to 7.9 hours of sleep. The ratio for women was 1.48. Men and women who reported sleeping ten hours or more had about 1.8 times the mortality of those who reported 7.0 to 7.9 hours of sleep. Those who reported using sleeping pills "often" had 1.5 times the mortality of those who "never" used sleeping pills. These results do not prove that mortality could be reduced by altering sleep durations or by reducing hypnotic prescribing. Rather, studies are needed to determine the causes of these mortality risk factors.