OBJECTIVE: To assess risk factors for legal induced abortion related deaths. METHODS: This is a descriptive epidemiologic study of women dying of complications of induced abortions. Numerator data are from the Abortion Mortality Surveillance System. Denominator data are from the Abortion Surveillance System, which monitors the number and characteristics of women who have legal induced abortions in the United States. Risk factors examined include age of the woman, gestational length of pregnancy at the time of termination, race, and procedure. Main outcome measures include crude, adjusted, and risk factor-specific mortality rates. RESULTS: During 1988-1997, the overall death rate for women obtaining legally induced abortions was 0.7 per 100,000 legal induced abortions. The risk of death increased exponentially by 38%for each additional week of gestation. Compared with women whose abortions were performed at or before 8 weeks of gestation, women whose abortions were performed in the second trimester were significantly more likely to die of abortion-related causes. The relative risk (unadjusted) of abortion-related mortality was 14.7 at 13-15 weeks of gestation (95%confidence interval [CI] 6.2, 34.7), 29.5 at 16-20 weeks (95%CI 12.9, 67.4), and 76.6 at or after 21 weeks (95%CI 32.5, 180.8). Up to 87%of deaths in women who chose to terminate their pregnancies after 8 weeks of gestation may have been avoidable if these women had accessed abortion services before 8 weeks of gestation. CONCLUSION: Although primary prevention of unintended pregnancy is optimal, among women who choose to terminate their pregnancies, increased access to surgical and nonsurgical abortion services may increase the proportion of abortions performed at lower-risk, early gestational ages and help farther decrease deaths. OBJECTIVE: To assess risk factors for legal induced abortion related deaths. METHODS: This is a descriptive epidemiologic study of women dying of complications of induced abortions. Numerator data are from the Abortion Mortality Surveillance System. monitors the number and characteristics of women who have legal induced abortions in the United States. Risk factors include age of the woman, gestational length of pregnancy at the time of termination, race, and procedure. risk factor-specific mortality rates. RESULTS: During the period from 1988 to 1997, the overall death rate for women obtaining legally induced abortions was 0.7 per 100,000 legal induced abortions. The risk of death increased exponentially by 38% for each additional week of gestation. women whose abortions were performed at or before 8 weeks of gestation, women whose abortions were performed in the seco nd trimester were significantly more likely to die of abortion-related causes. The relative risk (unadjusted) of abortion-related mortality was 14.7 at 13-15 weeks of gestation (95% confidence interval [CI] 6.2, 34.7), 29.5 at 16 -20% (95% CI 12.9, 67.4), and 76.6 at or after 21 weeks (95% CI 32.5, 180.8). Up to 87% of deaths in women who chose to terminate their pregnancies after 8 weeks of gestation may have been avoidable if these women had accessed abortion services before 8 weeks of gestation. CONCLUSION: Although primary prevention of unintended pregnancy is optimal, among women who choose to terminate their pregnancies, increased access to surgical and nonsurgical abortion services may increase the proportion of abortions performed at lower-risk, early gestational ages and help farther decrease deaths.