A postal survey was undertaken between 1985 and 1987 to assess the changes that have taken place in fetal monitoring practice since 1977. Replies were received from 253 (92%) of 276 consultant obstetric units canvassed. Biophysical methods of assessing fetal wellbeing in the antepartum period are almost universally employed and are accepted as the best discriminators of the need to deliver the pregnancy. The number of fetal heart rate monitors on labour wards has increased by 88%. Overall, 63% of units monitor more than 60% of their patients in labour, and 87% permit suitably trained midwives to apply fetal scalp electrodes, which must now be regarded as standard practice. There is still a need for a simplified technique for fetal blood sampling. The perinatal mortality rate correlates with a premuturity factor and probably bears a greater relation t o the population served than to the degree of monitoring provided.In 1978 a postal survey of intrapartum fetal monitoring practice in the UK (Gillmer & Coombe 1979) received replies from 244 (92%) of the 264 units canvassed. The approximate percentage of patients monitored electronically during labour in these units displayed a bimodal frequency distribution with peaks at 2C-30 and 8@-90%. This was interpreted as reflecting a division of opinion about the need to monitor all or only 'high-risk' patients. At that time, however, nearly 25% of units were unable to monitor all their high-risk labours. Only half the units allowed midwives to apply fetal scalp electrodes, and only 40% used fetal blood pH estimation. It