In order to gain a better understanding of the dynamic characteristics of forest landscapes and provide scientific bases for forest landscape restoration and its sustainable development at the community level, we analyzed dynamic characteristics and effect factors of forest landscapes from 1994 to 2014 using the methods of landscape indices and logistic regression analysis. The process was based on the use of high-resolution SPOT-5 images, TM images, household survey data, and some related auxiliary data. Shixing village, Shizhu County, located in the Three Gorges Reservoir Area, was selected as the forest sampling site. The results revealed the following: Forest landscape has been the main landscape type in the sampling area for the past two decades. The degraded primary forest mainly distributed in hills and low hills in the central and east part of the study site, while the secondary forest, degraded forest land, and plantations, which were distributed in the matrix of the major landscape, showed a mosaic pattern. The forest dynamics characteristics in these two decades reflected that the forest area reduced first, and followed by an increase, thus constituting a U-shaped trend. The dynamic evolution of the landscape in the period from 1994 to 2004 showed an inverse process to the period from 2004 to 2014. This process occurred mainly in the contiguous areas between forest landscape and paddy fields, dryland, human settlements, and roads. In terms of the spatial pattern, the recovery area in the period of 2004-2014 was almost the same as the degraded area in the period of 1994-2004. At patch level, the forest landscape was significantly heterogeneously distributed in the study site and has been for the past 20 years. The total ranking of forest landscape fragmentation in three different years was degraded forest land > plantation > secondary forest > degraded primary forest. In addition, the degraded primary forest, secondary forest, and degraded forest land had the same fragmentation tendency of arising first and reducing later. At the landscape level, the forest landscape fragmentation tendency was the same as that at the patch level, and the degree of fragmentation in 2014 was lower than that in 1994, which indicated that the speed of ecological integrity recovery was slower than the speed of degradation. Socio-economic factors had a significant influence on forest dynamics from 1994 to 2014. First, the most important factor was the change of residential population density. The fluctuation of the population density directly affected ecological pressure and led to the degradation and recovery of the forest landscape. The second critical factor was the change in the net income of farmers, especially the increase from 2004 to 2014, which resulted in the transformation of the energy structure and the reduction of dependence and interference in the forest landscape. The spatial distance factor had a steady effect on forest landscape dynamics, as human disturbance factors affected forest landscape via accessibility based on spatial distance. Overall, natural factors control the entire landscape pattern of the forest, but only slightly affected forest landscape in the short term. Secondary forest and degraded forest land were influenced by the natural environment owing to their location in the developing suburban area.