To explore the multimodality of two representative EFL textbook series for Chinese college students, their visual and verbal semiotic modes were compared. The target textbooks are "Experiencing English" and "New Century College English". Through multimodal discourse analysis, the study aims to shed some light on how to develop high-quality multimodal EFL textbooks. The main findings are: (1) "EE" and "NCCE" are similar in the fact that their representative multimodal texts are visually-verbally coherent and both demonstrate prominent features for intersemiotic semantic relations; (2) their differences are "EE" displays a higher degree of interpersonal intersemiotic complementarity and multimodality facilitates the realization of different modern educational concepts--constructivism in "EE" and humanism in "NCCE"; and (3) such differences are related to or may partially result from the differences in the language difficulty of textbooks and English proficiency of target learners. As a pioneering attempt to probe into the possible relationship between multimodality and modern educational concepts in EFL textbooks, the study shows the importance of properly arranging the different modes in a double-page spread. It also suggests that EFL textbook compilers consider the learners' English proficiency and appropriately employ the variety and number of multimodal resources to achieve optimal intersemiotic complementarity.