Hanging nest-boxes were used to study the nest predation of Great Tits in secondary broad-leaved forest from year 2004 to 2008. The results showed a significant difference between nest-boxes of different age groups with a chi-square value of 34.55 (P<0.01, df=3). The predation rates were the lowest in the new nest-boxes and the highest in the second-year nest-boxes. There was a large decline in predation rate in the third-year boxes and a slight climb-up again in the fourth-year boxes. This is consistent with the hypothesis of the nest predation rate increases as the age of nest-box increases, as a result of long-term spatial memory of the predator. To verify the hypothesis that Great Tits could adjust the clutch size after their nests are predated, we conducted a t-test to compare the clutch size of predated nests to the non-predated nests. With a t-value of 3.6488, the clutch size of preyed nests was significantly smaller than non-preyed nests (P<0.01, df=16). Habitat factor analysis (PCA) of prey nests revealed that the height of the nest-boxes and the degree of nest site coverage were the main habitat factors impacting the nest predation, followed by the density and height of shrubs as secondary impacting factors.