The pregnancy outcomes on cases of Down syndrome diagnosed prenatally in which the mother did not elect termination were evaluated in data reported to a comprehensive Register of Down syndrome for England and Wales for 1989-94. In the 168 cases in which placental biopsy was not used, the overall rate of spontaneous loss was 35%, but this figure masks considerable heterogeneity by gestational stage at ascertainment. Data on ages at diagnostic procedure and on pregnancy termination enabled a more precise survival analysis. The loss rates were approximately 50% for those fetuses ascertained at 15-17 completed wk, 43% at 18 wk, 31% at 19 wk, 25% at 20 wk, and then a leveling off at approximately 20%-25% for fetuses ascertained at 21-28 completed wk. For fetuses ascertained prior to 18 wk, there was no evidence that maternal age was associated with fetal loss, consistent with earlier reports. At 18 wk and after, however, maternal age was on the average approximately 3 years greater in fetuses that were lost. Comparison of successive gestational birth cohorts provided no evidence in these 168 cases that the diagnostic procedure itself had any effect on loss or that selective ascertainment of mothers in risk of loss had any effect on the results. In contrast, in the 21 cases in which placental biopsy had been undertaken, the overall loss rates were not only higher when appropriate comparisons could be made, but there was some evidence for selective ascertainment and/or procedure-associated losses.(ABSTRACT TRUNCATED AT 250 WORDS)