The article considers a problem of classification of text documents of a higher educational institution using machine learning methods such as the Bayes method, the k-nearest neighbors algorithm, the decision tree method, and the support-vector machines. The study is conducted on a set of documents of the Siberian State Automobile and Highway University, Omsk, using the Python programming language. Preliminary processing of documents for the study is carried out. All studied documents are divided into four classes namely the order, the instruction, the letter, and the notice of a vacant position.The process of classification of documents of a higher educational institution is presented. A comparative analysis of the classification results of each machine learning method is carried out on such metrics as the correctness of the classification algorithm, accuracy, completeness, F-measure, and the runningtime of the algorithm. As a result of the research, the author gives recommendations on the application of the considered methods for the classification of university documents. The author suggests using a combination of methods to optimize the operation of the documents classifier.