Abstract: The thesis is written to find out the occurrence and epidemic regularity of wheat stripe rust in Nanchong area, which will do a good job for monitoring, early warning and comprehensive of wheat stripe rust in time and accurately, meanwhile reduce the high air spread of stripe rust pathogen quantity that coming to Sichuan southeast spring epidemic area, Chongqing, Hubei, Hunan and other neighboring region, main wheat production district in east China. Thus, the wheat production, trade security and sustainable development can be ensured. Over these years, by the methods of systematical monitoring, regular surveys, field investigation, stripe rust resistance identification, physiological sampling inspection of morbidity plant, comprehensive treatment and meteorological data analysis, we have firstly discovered the initial bacteria source of wheat stripe rust, propagation path, afferent?peak?time, local epidemic peak time and the relationship between epidemic peak time, afferent time and afferent peak time. This study shows that the earlier the symptoms occurred, the more serious it will be. There are two afferent?infection?peak time. The first one plays a key role in bacteria spreading which existed in the field for 1~3 epidemic peaks and the area influenced by the first epidemic peak will play a decisive role in the same year. The cumulative occurrence area in late January is highly in correlation to the annual occurrence area with the correlation coefficient hitting 0.7692. In recent 16 years, the proportion of serious occurrence is 81.25% while light occurrence is only 6.3%. The number shows that the stripe rust is in an epidemic period. The internal reason is the reduction or loss of wheat variety’s resistance to tripe rust for a new physiological race of rust is becoming pathogenic stronger and be the major race. If there are big fluctuation of temperatures in warm winter and spring, or fog and dew days slants much, they would directly cause stripe rust in serious occur area, frequent occur area, early occur area and windy outlet valley area. Since 2002, by analysing the initial period, twice afferent range of foreign bacteria source, 3 epidemic peak times of occurrence area and the relation coefficient, we combine the field monitoring data and meteorological forecast to make a long, medium and short-term prediction. The short-term prediction accuracy rate of wheat stripe rust is 100% in every year while the medium and long-term prediction accuracy rate reaching more than 98% and 95% respectively, which increased by 5~15 percentage points than those before 1998.