Based on the element diffusion theory and analysis to the experimental conditions, in this paper， it is concluded that the “closure temperature of U Pb isotope system in zircon >900℃” drawn from experiment of U Pb diffusion in zircon by heating zircon crystals done by Lee et al.(1997) and Cherniak et al. ( 2000) may be only applied to understanding and explaining the behaviour of U Pb system of inherited zircons within granite during the heating and partial melting process in source region of the granite, but not be indicated that the zircon crystallized from the granitic magma has the closure temperature of U Pb isotope system as ≥850℃. The frequency analysis of 523 coupled age differences Δt (Δt = tZr-tRb ) between the zircon U Pb age (tZr) and whole rock Rb Sr age (tRb) for granite plutons shows that ① the Δt histogram displays a symmetrical normal distribution (skewness CSK=-0.205; kutrocess CKU=5.093), the Δt are both positive and negative; ② the distribution of Δt not possess eitheir positive deviation or negative deviation and are controlled by accidental factors, i.e. experimental error. Using least squares regression procedure, a best regression equation (tZr =0.999775×tRb +0.06898 Ma) with high correlation coefficient (R=0.997) and regression coefficient closed in 1 (a=0.999775) is fitted to the 523 pairs of zircon U Pb age and whole rock Rb Sr age of granite plutons. These statistical characteristics indicate that for granites as a whole both zircon U Pb dating age and whole rock Rb Sr isochron age are consistent within error.