The anomaly of increasing temperature which can show in thermal infrared images of satellite before earthquakes relates to the gases release from the crust. Because the gases such as CO2 and CH4 were monitored or sampled only on a few sites, scope of the gases release before earthquakes were not understood enough. From MOPITT data of America EOS satellite, we found a largescale discharge of CO from the Tibetan Plateau on April 30, 2000 (before it, CO continuously released at least for some days). The image shows that the area with abnormally higher CO content bears circle structure with irregular forms, its accumulative length and area respectively were about 3200 km and 2.67×106 km2, and volume percentage of CO is higher in inner circles than that in outer ones. The region with the largest CO volume percentage (31×10-8≥φ(CO)＞27×10-8) distributes roughly in EW direction with the length about 800km, width about 280 km, and area about 22.41×104km2. [WTBX]φ[WTBZ](CO) in the anomaly area was about 1.57~4.10 times the normal value observed in January, 2002. This phenomenon has a good coherence with large-scale temperature increase on several sites of the Tibetan Plateau from April 29 to 30, 2000. All of them are precursors of the earthquake with Ms 5.9 at Jingtai, Gansu province, June 6, 2000 and that with Ms 6.9 in the north of Burma, June 8, 2000. This means that gases-hot mechanism of earthquake precursors is truthfulness, and low value center of ozone over Tibetan plateau may relate to oxidation of CO in the upper air.