This article examines the variation in internal and external sandhi imposed upon two succeeding words in a seemingly identical phonetic context in Vedic Sanskrit. This variation has been considered to be caused by the syntactic relation between these words, with internal sandhi occurring where the words show a closer syntactic relation, and external sandhi occurring where they do not show such a relation. Since this claim has not been substantiated sufficiently, the present study seeks to test it, offering for the first time an exhaustive morphosyntactic and prosodic survey of one pair of such sandhis. The findings of this survey indicate that syntax plays no role in the choice of sandhi, but that the variation is caused purely prosodically: external sandhi is the default, while in-ternal sandhi marks prosodic boundaries either preceding or succeeding the sequence of words displaying it.