ICESat-2 satellite data are collected by a multi-beam photon-counting system, which greatly improves data coverage density and data accuracy. The forward-scattering effect caused by clouds and aerosols in the atmosphere significantly influences data accuracy. In this paper, an apparent surface reflectance-based cloud-detection algorithm using the pulse-folding characteristics of the ATLAS photon-counting system is proposed. The ICESat-2 cloud-detection process and its related algorithm are described in detail, which will be helpful for the application and expansion of ICESat-2 data in studies of the atmosphere. The results will provide reference that will be of significance for the correction of subsequent atmospheric effects in laser altimetry acquired by Chinese satellites.