This paper proposes an automated methodology for extracting fetal heart rate from single cutaneous potential abdominal signal. The proposed method calculates fetal heart rate in three steps. First, two-tier method based on singular value decomposition and polynomial classifiers is applied to the abdominal signal to separate the fetal ECG and the maternal ECG. Second, the fetal R-peaks are detected from the separated fetal ECG signal. Third, false R-peaks are eliminated and fetal heart rate is calculated. The proposed method is validated on both synthetic and real abdominal ECG and yielded a heart rate accuracy of 96%. The proposed method is advantageous since it requires only one lead in contrast to other methods which require multiple leads. © 2008 IEEE.