In the downstream of SOFC systems, higher oxygen partial pressure may cause oxidation-induced Ni anode degradation, associated with the formation of NiO and/or Ni(OH)2. Under experimental conditions for Ni anodes exposing to high oxygen partial pressures, conduction pathways could be destroyed resulting in cell performance degradation. This study is therefore focusing on the changes in microstructure and cell performance at high fuel utilizations. The objective of this study is to derive guidelines and limit for high fuel utilization operation towards higher power generation efficiency. The cells used consist of Ni-based anode with scandia-stabilized zirconia (ScSZ), ScSZ electrolyte, and lanthanum-strontium-manganite(LSM)-ScSZ cathode. Reference electrode made from Pt paste was attached to the cathode side to separate anode and cathode overvoltages. The electrochemical properties of the cells were measured at 800oC by feeding humidified fuel, H2 (20 cc/min)-H2O (80 cc/min)-N2 (100 cc/min), to the anode, and air (150 cc/min) to the cathode. As carrier gas, N2was fed to lower the humidifier temperature for simulating higher humidity conditions. The condition for Ni oxidation was estimated by thermochemical equilibrium calculation software, HSC Chemistry. Performance stability was examined at a constant cell voltage for a constant partial oxygen pressure at the anode for 100h. I-V characteristics before and after the 100h test were measured for confirming performance changes. Focused-ion-beam-scanning electron microscopy (FIB-SEM) observation was conducted for the anode, reconstructed by the brightness differences of BSE images, and microstructures were observed in detail by STEM. Anode voltage threshold was derived from the oxygen partial pressure at the boundary where both Ni and NiO coexist in the phase stability diagram as shown in Fig. 1. The boundary shifted to higher oxygen partial pressure as temperature became higher. The threshold voltage was 0.701 V at 800oC. The performance stability was examined for 100h at the constant cell voltage of 0.6 V, showing stable performance. The resistances did not change before and after the test, and no microstructural change of the anode was observed. The performance stability was also examined at the constant cell voltage of 0.5 V in order to set the experimental condition near the threshold. Anode voltage started to vibrate during the test when anode voltage approached about 0.7 V. The anode-side ohmic and nonohmic resistances increased during the vibration, suggesting Ni oxidation. But the electrochemical measurements revealed no change in anode resistances, indicating that the oxidation of Ni was limited to their surfaces during power generation. The number of isolated Ni particles was still small judging from FIB-SEM reconstruction images, maintaining their conduction pathways. The presence of NiO was also not confirmed by STEM mapping images. The performance stability was also examined at the constant cell voltage of 0.65 V when further humidified fuel, H2 (5 cc/min)-H2O (95 cc/min)-N2 (100 cc/min), was supplied. The performance was stable as shown in Fig. 2 and it was verified that power generation was possible even under such a condition if the anode voltage was kept sufficiently higher than the threshold. Figure 1