OBJECTIVE To compare spectrophotometric saturation values of fetal blood sampling to the saturation readings by pulse oximetry. METHODS During a clinical trial, fetal oxygen saturation was monitored during labor by a fetal oxisensor and a fetal pulse oximeter. Fifty-one fetal scalp blood samples were assessed because of abnormal fetal heart rate (FHR) patterns. The pulse oximeter displayed only signal quality readings. The investigator had to perform adjustments if signal quality fell below 50%. After delivery, the saturation at the moment of fetal blood analysis could be read from a printout and compared to the saturation values of scalp blood sampling. RESULTS The share of usable signal time was 51% overall, but only 40% in the 20-minute period during fetal blood sampling. Comparison with the reference method resulted in a median deviation of 6% (tenth percentile -10%; 90th percentile 18%) for pulse oximetry. The correlation coefficient between saturation values by pulse oximetry and fetal scalp blood sampling was 0.67. The correlation coefficient with the partial pressure of oxygen and oxygen saturation by pulse oximetry was 0.61, whereas it was 0.88 between partial pressure and saturation from the spectrophotometric analysis of the scalp sample. CONCLUSIONS Fetal pulse oximetry corresponds satisfactorily to results from fetal blood analysis. Low invasiveness and continuous monitoring are the advantages of this method. At present, the available sensor generates only a limited amount of signal time. However, in combination with FHR monitoring, pulse oximetry promises greatly improved detection of fetal hypoxia.