A retrospective comparison of cytogenetic and ultrasound findings in first trimester spontaneous fetal loss after demonstration of cardiac activity was made. The crown-rump length (CRL) was measured twice for each fetus resulting in spontaneous abortion: (i) CRL was measured in the viable state while demonstrating cardiac activity, and the growth deviation was expressed as the measured/expected CRL ratio (M/E CRL ratio); (ii) in the same fetus, CRL was measured after confirmation of fetal death, and designated as the post-mortem CRL. The chorionic tissues of these abortuses were karyotyped. The CRL of fetuses which resulted in normal deliveries were also measured as controls. As a result, 16 of 24 abortuses displayed an abnormal chromosomal analysis (67%). The mean M/E CRL ratio of still-viable fetuses was smaller than that of control fetuses (0.74 +/- 0.20 versus 0.98 +/- 0.13 respectively, P < 0.01). The differences in ratio between karyotypically normal and abnormal abortuses were not statistically significant. The post-mortem CRL of dead fetuses was > 20 mm in four of five monosomy X, two of three 21-trisomy, one of three triploidy and none of eight embryos with normal karyotype and five other trisomies. In conclusion, our study demonstrated that the M/E CRL ratio could be used as a predictor of spontaneous abortions, although it does not discriminate abnormal karyotypes from normal ones. The embryos with a post-mortem CRL more than 20 mm have a higher likelihood of suffering monosomy X or 21-trisomy. The ultrasonographic findings might offer a cytogenetic clue as to a possible cause to the developmental arrest.