The article explores two media discourse formats, film and TV interview as manifesting the specific distribution of two communicative modalities — speech and gesture — in the multimodal deixis of discourse space construal in narratives. To identify the attribution of speech and gesture to two media formats, the study contrasts the plot contents and viewer engagement as two major coordinates of discourse space in media. We hypothesize that since these two formats display differences in their plot content and viewer engagement characteristics, their distribution might affect the way the actors employ speech and gesture to present the discourse space. The research data are 20 film and TV interview narratives with 10 highly rated Russian actors (5 men and 5 women). The study exploits the method of multimodal analysis to determine the contingency patterns of discursive schemata presenting discourse space in speech and of discursive schemata presenting object of reference, move and frame of reference in deictic gesture. The results show that multimodal deixis contributes to intensifying both types of these characteristics. While multimodal deixis in film manifests the preference for metonymic pointing, placement indication and speech event construal; in TV interview, we observe the significant prevalence of metaphoric pointing, path indication and narrative event construal. Additionally, TV interviews display the speech schemata specifying information and accentuating the place of referent within the speech or narrative event. Overall, the study shows that there are two basic types of discourse characteristics which produce higher distinctions in multimodal deixis between film and TV interview formats; these are different narrative regimes and different combination of staged performance and spontaneity.