In Danish, as in other languages, prosody assignment is fairly well described as a function of lexical and syntactic structure. So in principle, prosodic clue assignment should be open to machine learning techniques. This paper presents an experiment using transformationbased ML for unsupervised learning of Danish main stress assignment. The trained stress assigner is compared to the leading Danish text-tospeech system. In conclusion, ML for prosody assignment is advocated as an attractive alternative to naive word mapping as well as to labourintensive grammar writing.