Abstract This article introduces micro‐prosody as the study of the duration and timing of speech events. We present a descriptive framework, formalising micro‐prosody in terms of gestural landmarks and coordination relations between them, and we use the framework to illustrate different patterns of micro‐prosody across languages. We show that potential ambiguity between coordination relations can be resolved by considering how they structure natural variation in speech. The framework presented here is intended to offer effective tools for phonetic documentation. We end with some considerations for a theory of micro‐prosody, including how micro‐prosody relates to other levels of phonological structure, and a brief discussion of different data types that can be used to infer aspects of micro‐prosody.