Recently, we have shown that visual speech can recalibrate auditory speech identification. When an ambiguous sound intermediate between /aba/ and /ada/ was dubbed onto a face articulating /aba/ (or /ada/), then the proportion of /aba/ responses increased in subsequent unimodal auditory sound identification trials. In contrast, when an unambiguous /aba/ sound was dubbed onto the face articulating /aba/, then the proportion of /aba/ responses decreased, revealing selective adaptation. Here we show that recalibration and selective adaptation not only differ in the direction of their after-effects, but also that they dissipate at a different rates, confirming that the effects are caused by different brain mechanisms.