The aim of the study was to investigate aspects of the epidemiology of Down's syndrome (DS) in a population subjected to continuous surveillance. Data were obtained from the Glasgow Register of Congenital Anomalies. Over the period 1980-90, 173 cases of DS were identified (1.2 per 1000 births), of which 135 (78.1%) were livebirths, five (2.9%) were stillbirths and 33 (19.1%) were terminations following prenatal diagnosis. The increasing risk of DS with advancing maternal age was confirmed, although most DS pregnancies occurred in women aged under 35. While the birth prevalence of DS did not vary significantly over time, there was a significant increase in DS pregnancy prevalence. The proportion of all DS pregnancies terminated showed a statistically significant rise from 14% in 1980-87 to 31% in 1988-90 following the introduction of a new screening marker for DS risk. Almost half of DS pregnancies in women aged 35 and over ended in termination following prenatal diagnosis compared with only 7% in women under 35. There was a statistically significant excess risk of DS pregnancy in the more prosperous areas of the city; this excess persisted, but was not statistically significant, after standardising the rates for maternal age. These findings suggest that there is no scope for reducing service provision for DS children on the basis of the epidemiological trends, that the antenatal screening programme, while benefiting from the recent introduction of serum markers of risk, continues to have a greater impact on older women, and that further aetiological research using population based registries is required.