Abstract: In this paper, common criterions about residual strengthevaluation at home and abroad are generalized and seven methods areacquired, namely ASME-B31G, DM, Wes-2805-97,CVDA-84 and J integral methods.BP neural network are combined with genetic algorithm (GA) named by modifiedBP-GA methods to successfully predict residual strength and criticalpressure of injecting corrosion pipelines. Examples are shown thatcalculation results of every kind of method have great difference andcalculating values of Wes-2805-97 criterion, ASME-B31G criterion and CVDA-84criterion fracture mechanics model are conservative and higher than those ofJ integral methods, while calculating values of DM fracture mechanics modelare dangerous and less than those of J integral methods and calculatingvalues of modified BP-GA methods are close and moderate to those of J integral methods. Therefore modified BP-GA methods and J integral methodsare considered better methods to calculate residual strength and criticalpressure of injecting corrosion pipelines.