OBJECTIVE: The aim of this study was to assess the risk of fetal loss attributable to cordocentesis at midgestation. STUDY DESIGN: A cohort study was conducted during the period 1989-1999. Women undergoing cordocentesis between 16 and 24 weeks' gestation with singleton pregnancies without obvious fetal anomaly were recruited into the study group. The control subjects were selected prospectively on a one-to-one basis with strict matching for maternal age, parity, gestational age at recruitment, and socioeconomic status. Both groups were prospectively followed up until delivery. RESULTS: A total of 1281 women with successful cordocentesis and their matched control subjects were recruited to the study. After exclusion of some pairs because of loss to follow-up or fetal malformations or severe disease necessitating termination of pregnancy, 1020 matched pairs were available to be compared with respect to fetal loss rate and pregnancy outcomes. The fetal loss rate was significantly higher among the study group (3.2% vs 1.8%; P < .05, McNemar test). However, there were no significant differences in other obstetric complications between the study and control groups. CONCLUSION: The incremental fetal loss rate associated with cordocentesis at midgestation was about 1.4%.