Electrocardiogram (ECG) is an important biomedical signal which shows the electrical activity of heart. ECG waveform provides the valuable information for detection of abnormal heart diseases. A VLSI based modified arithmetic discrete wavelet transform for arrhythmia detection and its FPGA implementation. Time and frequency variations of QRS complex in ECG signal are identified. Then comparing this with normal ECG and identifying whether the patient is having arrhythmia or not. The modified discrete wavelet transform is used for compressing the ECG signal so it only consumes area of 6% and faster than the other approaches.