As decreased foetal movement (FM) may indicate impaired foetal health, FM recording has been suggested as a method of assessing foetal well-being. A non-intrusive, automated method of recording FM (FM-detector), was compared with maternal and ultrasonographic assessment of FMs in 24 women in the third trimester of pregnancy. The FM-detector detected a greater proportion of ultrasonographically recorded FMs than the mothers did (median 70% and 38%, respectively; p less than 0.001). Parity, gestational age, placental site or thickness, maternal weight or the distance from the maternal abdominal surface to the amniotic cavity did not affect the ability of the FM-detector to detect ultrasonographically recorded FMs. The estimation of FM strength by the FM-detector agreed fairly well with the assessment of FM strength by the ultrasound observer. The FM-detector would seem suitable for clinical use, as in the examination of pregnant women complaining of feeling 'less FM'.