An ECG signal processor (ESP) is proposed for the low energy wireless ambulatory arrhythmia monitoring system. The ECG processor mainly performs filtering, compression, classification and encryption. The data compression flow consisting of skeleton and modified Huffman coding is the essential function to reduce the transmission energy consumption and the memory capacity, which are the most energy consuming part. The classification flow performs the arrhythmia analysis to alert the abnormality. The proposed ESP IC is implemented in 0.18-microm CMOS process and integrated into the wireless arrhythmia monitoring sensor platform. By integration of the ESP, the total system energy reduction is evaluated by 95.6%.