This article proposes a method for charge accumulation in a large-pixel CCD, which includes two aspects: The first is the subdivision of vertical and horizontal pixel structures within the device and the sum of charge packets. The second is the multiphase nonuniform driving in driving timing. This method is applied to a 210 μm×210 μm pixel CCD and is used to build hardware circuits and design software. Its effectiveness is verified via acquisition imaging and parameter testing. Results show that a large full-well capacity and dynamic range are achieved without compromising transfer efficiency.