In this paper, we have investigated the effectiveness of Particle swarm optimization (PSO) with extended Kalman smoother (EKS) for fetal ECG extraction from single channel electrocardiogram (ECG) recorded at abdominal area of mother s skin. The abdominal ECG is considered to be composite as it contains both mother s and fetus ECG and is dominated by maternal ECG component. To extract the fetal ECG, the cancellation of maternal component and noise in the abdominal ECG is very much required. PSO is used for selection of optimized parameters which are required to model the maternal ECG component according to the abdominal ECG and also to initialize the parameters of EKS. EKS framework is used for extraction of fetal ECG by using single channel abdominal ECG. Our results demonstrate the effectiveness of the proposed technique in extraction of the fetal ECG signal from single channel abdominal signals by using 2 real abdominal ECG database namely DaISY database, non-invasive fetal ECG database. Our proposed method shows accuracy of 89.7%, sensitivity of 93.2%, and positive predictive value of 94.97% for fetal ECG extraction from noninvasive fetal ECG database.