Ca-accumulating activity of the fragmented freg sarcoplasmic reticulum was usually 160–200 in/imoles Ca/mg protein in its capacity and 3–5×106 M-2 in its binding constant. Occasionally, preparations with larger capacity, nearly 300 mμmoles Ca/ mg protein, which were more sensitive to caffeine, were obtained. The properties of frog microsomes, which is mainly composed of the fragmented sarcoplasmic reticulum, were essentially the same as those of rabbit microsomes, although some differences were noticed in their responses to temperature, ADP, or pH. In the presence of high concentrations of ATP, Ca-uptake was carried out in two steps in the presence of an ATP regenerating system. The capacity became maximal at 1 μM Mg-ATP and 1 mu Mg2+, whereas the rate at 30 μM Mg-ATP and 3 mM Mg2+. The minimum effective concentration of oxalate was 0.3–0.4 mM. The time course of Ca-uptake in the presence of oxalate consisted of two phases. The first phase, the rate of which was the tame as that in the absence of oxalate, was followed by the second phase with a constant rate determined by the oxalate concentration. The effect of Pi on the Ca-uptake was not simply ascribed to the formation of insoluble salts. Caffeine released Ca rapidly from fragmented frog sarcoplasmic reticulum. Ca-releasing action of caffeine was more effective in the heavier fraction (1, 200–7, 000Xff) than the lighter fraction (7, 000–54, 500×0), and at lower temperatures. The effect of caffeine was counteracted by procaine. These agreed with the results of Weber and Herz. In contrast to the result of Weber there was no difference in the effect of caffeine between 10 μM and 0.4 mM ATP. The effect of caffeine was dependent on the time when caffeine was added. Thymol showed essentially the same effect as that of caffeine, but was about thirty times as effective as caffeine. The mechanism of action of caffeine was discussed. It was considered to be the releaser of accumulated Ca rather than the inhibitor of Ca-uptake.