A tool developed in order to reduce the subjective aspect of archival classification and make it more consistent has been sought and demonstrated. Taking into account that misclassification may harm the vast majority of archival functions, especially evaluation and description, a software called Ontological Classifier (OntoClass) has been developed. That software, by creating an ontology from the classification plan of a document producing entity, is capable of determining the class to which a document belongs on the basis of authorised terms set out in a list. The theoretical rationale was carried out by means of a bibliographic search and the tool was developed using Python 3.7 programming language and SPARQL consultation language. From tests carried out with a simple ontology specifically created for this work, it appears that OntoClass achieves its purpose, despite the fact that testing in real-life situations is still needed and despite some requirements to be met in order to obtain positive results of its use.