Detection of cardiac arrhythmias is a dominant problem in automatic systems of evaluation of parameters of electrocardiosignal (ECS), e.g., Holter monitors. Cardiac arrhythmias are diagnostically important param eters of cardiovascular system disorders (cardiac pulse propagation disorders and disorders of pacemaker pulses).Diverse algorithms for arrhythmia recognition fall into three groups: 1) temporal ECS analysis; 2) frequency ECS analysis; 3) combined methods [2].Methods of temporal ECS analysis are most frequent ly used. These methods are easily implemented and pro vide real time effective ECS analysis for arrhythmias. Up grading of methods of temporal ECS analysis for arrhyth mias is being continued for increasing noise resistance.Most methods for arrhythmia detection are based on comparison of mean time RR mean with current time RR i with fixed or programmable threshold values. The thresh old values are selected from clinical experience or second ary diagnostic symptoms (ratio of neighboring RR inter vals on current mode) [1]. RR interval duration is deter mined by comparing QRS complex with threshold values. The disadvantages of the point methods are as follows.1. Necessity of isolation of reference points in each cardiac cycle for determining RR i .2. Extrasystole complexes and pulse interference decrease reliability of detection of arrhythmias. Extrasystole complexes and pulse interference may result in false reference points or omission of reference points.3. Dependence on ECS shape. Certain types of ECS shape may result in false reference points or omission of reference points.The method for arrhythmia detection suggested in this work is based on an integral approach to ECS rhythm analysis and integral criterion (signal energy in time win dow).This method is based on the following assumptions. First, potential energy of heart in steady state mode is invariable depending on blood volume, vascular resist ance, psychological state, etc. [3]. Second, according to practical cardiologists, the amplitude-temporal parame ters of ECS without pathological (extrasystolic) complex es are invariable in time. PQ and ST segments are insignif icantly changed; TP segment is changed most significant ly, the TP segment time characterizing the cardiac cycle time. Thus, the energy of each cardiac cycle within a given ECS is virtually invariable. ECS wave energy is variable: QRS complex energy is larger than P wave energy and larger than T wave energy. Therefore, QRS complex contributes most significantly in cardiac cycle energy. Energy of the ECS segments is virtually zero.Energy E of each wave is a sum of squared discrete readings of signal:where i are numbers of discrete readings of wave within the range i = 0...n, where n is the number of the last wave reading (ratio of wave time b to sampling period dt: n = b/dt).The wave energy is:where k is coefficient depending on wave shape and sam pling period; a is wave amplitude. ECS waves can be approximated by linear, quadrat ic, or sinusoidal functions [2]. For linear functi...