We have been carrying out experiments that involve the mapping of excitation spread in the isolated rat atrial preparation using multiple-site optical recording methods [1, 2]. When the tetanus stimulation was applied to the preparation, long-lasting rhythmical excitation with high frequency, which looks like tachycardia, occurred. We call this phenomenon as “tachycardia-like excitation” [3 – 5]. We found the chaotic excitation spread during this event. Adult rats were anesthetized by inhalation of ether, and the hearts were removed quickly. The isolated atrial preparations were stained with a fast voltage-sensitive merocyanine-rhodanine dye (NK2761). The preparation chamber was mounted on the stage of a microscope equipped with 16 X 16 or 12 X 12 -element photodiode array. Optical action potentials were recorded from 256 or 144 adjacent areas of the preparation simultaneously. Then, maps of excitation spread were made. From the spatio-temporal patterns of excitation spread on the map, the characteristics of “tachycardia-like excitation” are summarized as follows. 1. Tachycardia-like excitation is the generation of (or “transition to”) a new pattern of excitation spread evoked by perturbing the “normal” pattern. 2. During its generation, transitional unstable complex (so called “chaotic”) patterns of excitation spread, with blocked areas and ectopic foci, are observed. 3. After this phase, a new quasi-stable pattern appears. It is “quasi-stable” because the newly emerged pattern often returns to the original “normal” pattern spontaneously. 4. Event-to-event variations are always observed. Furthermore, preparation-to-preparation variations are also observed. During the event of tachycardia-like excitation, cycle-to-cycle variations (for circus movement) and/or site-to-site variations (for ectopic focus, i.e. spatial sift of the focus) are also observed. Based on the evidences shown above, we consider that tachycardia-like excitation is an example of a functional “self organizing system”, although the physiological background is still unclear. Especially, the event-to-event variations seem to result from physiologically trivial difference(s) of the initial conditions. This characteristic of a “complex system” strongly supports this idea. We consider the quasi-stable state observed here as the “physiology-specific attractor”. This study was carried out in accordance with Act on Welfare and Management of Animals of Japan, and approved by the Animal Care and Use Committee, Tokyo Medical and Dental University.