Noise removal from a non-stationary signal like Electrocardiogram (ECG) signal is a major problem. Moreover, nullifying different noises using different adaptive algorithms –Least Mean Square (LMS), Normalized Least Mean Square (NLMS) etc. from ECG signal is one of the advance studies in biomedical signal processing. In this paper, we will discuss how we can remove different type of noises like 50Hz Power-Line Interference, Base-line Wandering and Muscle Contraction noise from an ECG signal using an adaptive filter. Different performance parameters such as, Signal-to-Noise Ratio, Mean Square Error and Root Mean Square Error are also calculated to compare the results. Real time data has been collected from MITBIH arrhythmia database. At the end, results show the better performance of adaptive NLMS filter for removing different noises over adaptive LMS filter.