This third part of the paper deals with the study of the relationships between fetal movements, fetal heart rate accelerations associated with such movements, fetal heart rate instability and neonatal outcome. No correlations has been found between absence of fetal movements and neonatal distress. A correlation has been found between the lack of fetal heart rate accelerations, the flatness of the record and poor neonatal outcome. In extreme situations (i.e. flatness in less than 10% of the record or in more than 80%) the presence or absence of accelerations does not add further useful information. Such information, however, is gained in the intermediary situations (the 'combined' recordings) and particularly when the record is between 51 and 80% flat where there appears to be an 85.6% risk to the fetus. Consequently, when trying to analyse an antenatal record it seems advisable to take primarily into account the percentage of flat recordings (providing the records are numerous enough and of sufficient length). Then, in records between 10 and 50% flat, the presence or the lack of spontaneous decelerations requires consideration whereas, when the record is between 51 and 80% flat, it is the presence or absence of fetal heart rate accelerations which is important.