Each mode of Audiovisual Translation (AVT) requires a slightly alternate approach. Whether a given audiovisual material is translated by the means of subtitling, dubbing or voice over often results in a very different translation of the same Source Text (ST), which is not surprising taking into consideration various technical constraints and translation tendencies typical of a given AVT mode. At the same time, audiences approach a given AV material with a prior set of expectations, which often include AVT mode preferences. Moreover, viewers’ prior knowledge and fluency in Source Culture (SC) affects their perception of the end translation. According to Marchant et al. (2009), “the overt visual attention detected by an eye-tracker gives us a window on internal systems” what, in turn, may be crucial for translators to improve their performance. The presented article is an attempt to show how audience really watches AV materials and how various modes of AVT affect viewers’ reception of a given production. The main objective of the paper is to answer the question: How a mode of AVT affects audience’s reception of humorous audiovisual materials deeply rooted in Source Culture? The analysis is based on the research involving eye-tracking devices conducted in May-July 2016 in Warsaw (Poland), during which the participants were asked to watch five short fragments of various humorous American productions deeply rooted in American culture (animated tv series: The Simpsons and South Park; animated feature film: Madagascar; tv series with actors: Gilmore Girls).