The Balinese are one of the ethnic groups of Indonesia, the majority of which are on the island of Bali, the language used is Balinese with three levels of sor-singgih (tigang soroh) guidelines, namely Basa Kasar, Basa Madia and Basa Alus. Balinese language also has the additions of pangater, seselan and pangiring. To facilitate the search for basic words in Balinese, a stemming process is needed. Stemming is the process of mapping and decomposing the form of a word into its basic form. The stemming process is very important in the information retrieval system process. In this study, the Balinese stemming process used the Rule Base Approach method. The data used in this study are 376 basic words in Balinese. This study aims to design an appropriate stemming application for Balinese stemming. The initial stage in the Balinese stemming process is to carry out the input process, preprocessing, filtering, case folding and tokenization. Each word is subjected to a stemming process to remove the additions of pangater, seselan, and pangiring. The results of the study indicate that the Rule Base Approach method can be used to stem Balinese texts, this can be seen from the results of the accuracy reaching 77.82%. Of course, in testing there are still failures caused by overstemming errors resulting from the stemming process.