Since it was possible for Ca²⁺, Mg²⁺-ATPase of sarcoplasmic reticulum (SR) to change its aggregation state in the membrane depending on temperature, and since the change could be the cause of the break in the Arrhenius plot of Ca²⁺, Mg²⁺-ATPase activity, the aggregation state of Ca²⁺, Mg²⁺-ATPase at 0°C in the membrane was compared with that at 35°C by freeze-fracture electron microscopy. These temperatures are below and above the break in the Arrhenius plot (about 18°C), respectively. Two kinds of samples were used; fragmented SR vesicles and egg PC-ATPase vesicles, a reconstituted preparation from purified Ca²⁺, Mg²⁺-ATPase and egg yolk phosphatidylcholine (egg PC). For both the appearance of particles in the fracture faces of the samples fixed at 0°C was similar to that at 35°C, and phase separation between protein and lipid was not observed even at 0°C. The size of the particles was measured and histograms of the sizes at 0°C and 35°C were made. The histogram at 0°C was similar to that at 35°C with a peak at 7.1nm, which is 1-2nm smaller than the value reported so far. The number of the particles per unit area of the membrane was also counted. The value at 0°C was similar to that at 35°C. These results indicate that Ca²⁺, Mg²⁺-ATPase of SR exists in the same aggregation state (estimated as oligomer based on the values obtained in this experiment) between 0°C and 35°C. Based on the results of this study we think that the break in the Arrhenius plot of Ca²⁺, Mg²⁺-ATPase activity in SR is not caused by the change in the aggregation state of Ca²⁺, Mg²⁺-ATPase.