Three questions in the analysis of Danish stød are addressed. (1) We try and establish an acoustic reality behind a proposed phonological analysis of consonants with stød as long (moraic) in opposition to short stødless (non-moraic) consonants and fail. (2) Vowels with stød are phonologically long; they are proposed also to be bi-moraic, with the second mora carrying the stød. We confirm the acoustic reality of length, but a moderate tendency only to acoustic bi-partition. (3) In a perceptual experiment vowels with stød were paired with long stødless vowels by half the subjects, whereas the other half paired syllables with stød, irrespective of vowel length. This is a good point of departure for an experiment to find out whether vowels with stød are cognitively bi-partite.