By applying linear trend method, HP filtering method and wavelet noise reduction method respectively，this paper makes comparison and analysis of such four dimensions as real-time, post-correction and quasi-real-time output gap estimation during 1986—2012 in China, the consistency from real-time to post-correction of the estimation from all kinds of methods, the consistency between inflation prediction capacity and historic economic volatility curve trends and the stability of the estimated results, concludes that the real-time output gap estimation result by wavelet noise reduction method is better, analyzes the influential factors of output gap from such two aspects as endogenous pull and exogenous drive, and the results show that the influence of technical progress volatility on GDP output gap is big in endogenous factors, the influence of international capital is second and the influence of human capital input is the smallest, meanwhile, the impact of the input of international capital and human capital on the later period is bigger than on the same period, and that in exogenous factors, the impact of urban per capita disposable income on output gap rate is big while output gap rate is negatively affected by population aging degree. Thus, output gap adjustment should start from technical progress promotion, human capital efficiency-increase, foreign capital investment attraction, income distribution optimization, population structure improvement and so on.