The maternal perception and the tocographic findings were quantitatively compared to the ultrasonic assessments of fetal movements. One hundred and eleven normal pregnant women (24-41 W) were examined with two ultrasonic real-time scanners and a cardiotocogram. At the strong fetal movements (body and extremities), the mother perceived 66% of fetal movements detected with real-time scanners, whereas 44% of fetal movements were detected with the tocotransducer. At weak fetal movements (movements of only fetal extremities), the mother perceived 23% of movements and with the tocotransducer 14% of movements were detected. On the other hand, when the mother perceived the fetal movements, 26% of which were false positive, whereas at the spikes of the tocographic findings, 44% were found to be false positive. These results demonstrate that for the clinical use, maternal perception is more reliable than tocographic findings. But for the NST of long duration, detection with tocotransducer is also useful.