Multimodal materials have become ubiquitous in language teaching classes. However, growing evidence suggests that misalignments between multimodal elements can mislead or even confuse students. To explore such concerns, this paper focuses on the design and pedagogical application of intersemiotically cohesive multimodal materials in an English for Specific Purposes (ESP) context. Specifically, utilizing a classroom-based intervention with two groups of Thai undergraduates (N = 41) enrolled on an English for agriculture course, we employed think-aloud protocols and video recordings to capture their interactions with multimodal materials we designed. The video recordings were then used for video-stimulated focus group interviews with six representatives from each group. This allowed us to explore how these students perceived and interacted with our designed materials and enabled us to investigate how a congruence between written text, visuals, hyperlinks to translations and audio, and consciousness-raising questions affected their learning experience. The findings revealed that well-integrated multimodal resources supported vocabulary acquisition and reading comprehension. However, students reported challenges with the abstraction of text and usability of hyperlinks, suggesting areas for further material refinement. Based on our results, we make recommendations for multimodal material design and underscore the critical need for cohesive educational resources to ensure learning effectiveness and accessibility.