Women’s gymnastics team competition is composed of high-low bars, balance beam, and vaulting horse and floor exercises these four items. In practical competitio ns, athletes strength surely decides team total sco re result, but a good starting line-up also has very important impac ts on competition results. The paper goes deeper in to exploring women’s gymnastics team’s team competitions forming problems, firstly according to known conditions ,i t provides each athlete score in the most pessimistic case, es tablishes objective functions to let total score to be as high as possible, combines with athlete participation rules , it gets constraint conditions, adopts 1~0 integer programming to establish model, applies lingo software to solve the model, and gets best starting line-up under pessimistic model; finally applies probability statistical theory and normal distribution knowledge to further deepen pro blem, by calculating, it gets variance ij D and expectation ji C , , and inputs them into standard normal programming functions so that can get objective functions, applies lingo software to solve best starting line-up.