The cytogenetic findings were analysed in a series of 500 pregnancies in which chorionic villi sampling was performed. In all cases a direct method was used, karyotyping being successful in 481 cases (96.2%). The main indication for sampling was maternal age over 36 (412 cases; 82.4%). Abnormal laboratory findings resulted in 24 terminations of pregnancy (4.8%); in addition five unexpected balanced chromosome rearrangements were detected. Twelve of 15 cytogenetic discrepancies were detected at amniocentesis, two after termination, and one at spontaneous abortion. Complete follow up data were available for the first 250 patients, among whom nine pregnancies (3.6%) ended in spontaneous abortion before the 20th week. There were no false negative findings. Seventy additional chromosome studies were performed because of failure of chorionic villi sampling or equivocal results, or for confirmation. Counselling before chorionic villi sampling should include the possibility that subsequent amniocentesis may be needed should mosaicism or other unexpected abnormalities be found. The success rate and accuracy of karyotyping chorionic villi samples by the direct method are acceptable but distinctly less than those of karyotyping cultured amniotic fluid cells.