OBJECTIVE The null hypothesis is that there is no difference in outcome when pregnancies with ultrasonographically documented subchorionic hemorrhages are compared with those without these hemorrhages. METHODS We performed a case-control study, utilizing our computerized ultrasonographic database. Cases were matched with two or three controls in two separate control groups. Matching criteria were maternal age, gestational age at scan, and invasive procedures (chorionic villus sampling or amniocentesis). General exclusion criteria were absence of fetal heart motion and fetal anomalies. Presence of subchorionic hemorrhage was an exclusion criterion for both control groups; however, in addition, presence of vaginal bleeding was a further criterion for one of the two. Statistical analysis was performed with chi2 analysis and Yates' correction. Odd ratios and 95% confidence intervals were calculated. RESULTS There was no difference in maternal characteristics between the cases and controls. The incidence of subchorionic hemorrhage was 1.3%. There was an increased risk of miscarriage (odds ratio 2.8, 95% confidence interval 1.7 to 7.4), stillbirth (4.5, 1.5 to 13.2), abruptio placentae (11.2, 2.7 to 46.4), and preterm labor (2.6, 1.5 to 4.6) when cases were compared with controls without subchorionic hemorrhage or bleeding. These risks were also increased in comparison with the control group with bleeding, except with respect to miscarriage. In this case the risks were similar in both cases and controls but increased with respect to the controls without bleeding. The mean birth weight was lower in the cases than in both control groups. CONCLUSIONS The presence of an ultransonographically detected subchorionic hemorrhage increases the risk of miscarriage, stillbirth, abruptio placentae, and preterm labor. The presence of bleeding alone appears to increase the risk of miscarriage. It is unclear whether the subchorionic hemorrhage is causative or whether it is simply a sign of an underlying process that produces these negative effects.