Tibetan sort will be widely used in every aspect of Tibetan language text information processing, including word, dictionary sequence, system software and other application software. This paper attempts to describe Tibetan writing stroke sorting rules which make more correct, reasonable induction, the purpose is to find a sorting algorithm model that could realize automatically the Tibetan strokes character. And which break the framework in Tibetan character syllable sequence depends only on the syllable key radical structure sorting traditional of thinking.