The effect of lumbar epidural block on placental intervillous blood flow (IBF) was estimated with the i.v. 133Xe method in 24 cases and in 14 controls. All parturients had uneventful pregnancies and were scheduled for elective caesarean section. Lumbar epidural block caused a significant decrease in maternal mean arterial pressure (MAP), when no preloading infusion with a plasma expander was given. In contrast, the MAP remained unchanged during lumbar epidural block when the parturients received a preloading infusion (Haemaccel, 100 ml/10 kg body weight within 10 min immediately before lumbar epidural block). The IBF decreased to some extent in the non-preloaded group, but increased temporarily in the preloaded group 15 min after the administration of lumbar epidural block. The mean changes of IBF in the non-preloaded and the preloaded groups differed statistically significantly from each other during lumbar epidural block. The IBF of the preloaded group did not differ significantly from the IBF of the control group receiving no lumbar epidural block.