Objective:To evaluate the clinical complications and diagnostic problems of chorionic villus sampling.Design:A pragmatic retrospective analysis.Setting:Tertiary obstetric referrals mostly to private practice; sampling carried out at Royal Prince Alfred Hospital; diagnostic analysis usually at Oliver Latham Laboratory, or the Clinical Immunology Research Centre, Royal Prince Alfred Hospital, New South Wales.Patients:1500 women in the first trimester of pregnancy requesting prenatal diagnosis because of a risk of chromosomal or inherited genetic disorder in the fetus.Interventions:Ultrasound‐guided passage of a catheter into the chorion through the cervix.Main outcome measures:Incidence of unintended abortion, preterm births, low weight infants and discrepant karyotypes.Results:There were 42 unintended abortions (3.0%), about 0.4% higher than the background abortion rate in women of similar age. Abortion did not occur more frequently in women with vaginal bleeding earlier in that pregnancy. Rates of preterm births and birth of low birthweight infants did not differ from the general population. A second diagnostic test was required in 68 women (4.5%). Mosaicism and tetraploidy were usually confined to the chorion.Conclusion:Chorionic villus sampling is an acceptable diagnostic test. Amniocentesis should be offered to patients who show mosaicism or tetraploidy.