The epidemiologic study of pelvic inflammatory disease (PID) is hindered by symptoms that differ substantially from case to case, and by a substantial number of "silent" cases with few or no symptoms. Demographic data indicate that PID is primarily a disease of the young. Race also appears to be a determinant, but whether for biologic or sociologic reasons is not known. Women who are not married face a greater risk for PID than married women. The majority of PID cases are associated with sexually transmitted diseases. Consistent with this finding is the pattern of risk factors related to sexual behavior: young age at first intercourse, high frequency of intercourse, and a large number of sexual partners all increase the risk of PID. On the other hand, barrier methods of contraception appear to reduce the risk of PID, as does oral contraception. Perhaps the best publicized risk factors for PID is another contraceptive, the intrauterine device (IUD). The increased risk of PID in IUD users appears to be confined to new users and women at high risk of sexually transmitted disease infection. Douching and cigarette smoking may also be associated with the development of PID, but the evidence is weak.