Gender representation has long been studied in both verbal and visual modes of ELT textbooks. However, regarding the visual mode, research has mainly focused on superficial analyses of how often each gender appears in different roles rather than on how the two genders are represented. The tools proposed in Kress and van Leeuwen’s (2006) social semiotics framework, however, permit deep analysis of images taking into consideration how pictorial elements are shown both alone and in relation to other pictorial elements, on the one hand, and the viewers on the other. Following the above-mentioned framework, the present study applied the three dimensions of representational, compositional and interactive meaning presented to 16 photographs randomly selected from the Interchange (Third Edition) series (Richards, 2005) to explore gender portrayals and disclose ideologies in the visual mode of the series. Qualitative data analysis showed some ideologies and stereotypical portrayals, each of which appeared either in one or a few photographs. Taken together, the findings indicated gender bias in favor of men in the series
