The Beserman dialect of the Udmurt language has a large number of clitics, mainly discourse markers, which have never been described in detail. These clitics occur very frequently in texts and often appear in clusters. The rules which govern the ordering of clitics within a cluster are especially interesting because the binary relation “clitic A normally precedes clitic B in a cluster” proves to be intransitive, thus making it impossible to use a clitic template to describe the ordering. In this paper, the rules which govern placement of individual clitics in a sentence and their mutual order in clusters are described. The research is based on a corpus of approximately 64,000 tokens and the authorʼs own field data collected in the village of Shamardan, Yukamensk region, Udmurtia.