A variety of land use prediction models can be observed elsewhere in the world, and they are doing a tremendous job in the planning its allied disciplines. But still it has been a distant factor in Sri Lanka. Unlike other models, cellular automata-based models are highly depending on the land cover data in past and present. But land cover data are developed by many government and private agencies with a range of objectives, and they are based on different classification schemes and have discrepancies (Mushtaq et al., 2022). Missing of a standard land use classification method has created a situation that all available land cover data and their applications have many limitations. Further, an analysis of cellular automata-based land use models shows that the accuracy increases as the heterogeneity of land use increases (PEI et al). Here, a standard land use classification method was developed considering many parameters such as availability, accuracy, informative, acceptability, validity and affordability. Land use legends from existing research articles, planning regulations and reports, project reports, published maps, satellite images, big data, etc., were thoroughly evaluated here. Further, a numerical analysis was carried out to showcase the existence level of each land use category at present.