The following article reports on a multimodal corpus study of English as if constructions. The results of this study suggest that formulaic and insubordinate as if constructions are prosodically chunked as clauses, with formulaic as if constructions uttered with significantly higher pitch and insubordinate as if constructions with lower pitch when being compared with subordinate uses. In addition, insubordinate as if clauses are occasionally accompanied by frowns. It is argued that, although both constructions convey an ironic interpretation, multimodal markers of irony play only a minor role in explaining the findings. Instead, it is argued that the non-verbal features are construction-specific and can reasonably be explained as cross-modal collostructions. As such, the present article provides a description of the non-verbal features accompanying English as if clauses and provides a theoretical explanation. In doing so, some modest evidence for a multimodal Utterance Construction Grammar is also presented.