Landscape pattern analysis is a popular topic in global change research, and appropriate grain size is the basis for accurate landscape pattern analyses. In this study, methods based on landscape metrics and spatial autocorrelation are used to determine the optimal grain size of the study area. Based on a grain size of 20 m, the landscape pattern of the core of the Zhengzhou and Kaifeng integration area was analyzed at class and landscape levels from 2005-15. The results show that over the study period, the landscape fragmentation of the study area increased by approximately 32.38%, the distribution of landscape types was homogenized gradually, and human impact was the main reason for the changes in landscape pattern. Additionally, cultivated land was the predominant landscape type, and the area percentage of cultivated land decreased from 79.01% to 60.01%. The patch number and total area of forests increased, the percentage of construction land increased by approximately 1.5 times, and the area and patch number of unused land gradually decreased. This research provides useful information for land use policy-making.