OBJECTIVE: To determine whether the prone sleep position was associated with an increased risk of the sudden infant death syndrome (SIDS). STUDY DESIGN: Population-based case-control study. PARTICIPANTS: Case subjects were infants who died of SIDS in King County, Washington. Control subjects were randomly selected infants born in King County. Up to four control subjects were matched on date of birth to each case subject. METHODS: During the study period, November 1992 through October 1994, sleep-position data were collected on infants who died of SIDS by the King Count Medical Examiner's Office during their investigation of the deaths. Parents of infants chosen as control subjects were contacted by telephone, and sleep position information was obtained. Infants who usually slept on their abdomen were classified as sleeping prone; those who usually slept on the side or back were categorized as sleeping nonprone. The adjusted odds ratio for prone sleep position as a risk factor for SIDS was calculated with conditional logistic regression after control for race, birth weight, maternal age, maternal marital status, household income, and maternal cigarette smoking during pregnancy. RESULTS: Sleep position data were collected on 47 infants with SIDS (77% of eligible infants) and 142 matched control subjects; 57.4% of infants who died of SIDS usually slept prone versus 24.6% of control subjects (p < 0.00001). The unadjusted odds ratio for prone sleep position as a risk factor for SIDS was 4.69 (95% confidence interval: 2.17, 10.17). After control for potentially confounding variables, the adjusted odds ratio for prone sleep position was 3.12 (95% confidence interval: 1.08, 9.03). CONCLUSION: Prone sleep position was significantly associated with an increased risk of SIDS among a group of American infants.