The field of language education has experienced a rise in using virtual reality (VR) to support interactive, contextualized, and collaborative language learning in recent years. The current study investigates the effects of auditory, visual, and textual input on speaking and writing in English for Tourism Purposes (ETP) through immersive, scenery-based virtual reality (SBVR) dyadic learning. Adopting a framework on ETP sociocultural, linguistic, and formulaic competence, the researchers designed relevant learning content using audios, images, and verbatim onscreen text on "EduVenture VR" to create an interactive, simulated tour. Thirty-eight English as a foreign language undergraduates participated in this single-group quasi-experiment. Various data sources included speaking and writing pre-, post- and delayed post-tests, a post-intervention survey, and reflections about the VR learning experience. The study reported significant speaking gains in language accuracy, destination promotion, knowledge retention, and replication of ETP language, as well as significant writing improvement in using sophisticated vocabulary, knowledge retention, and replication of ETP language. The study concludes that SBVR warrants positive impact on improving productive ETP skills. Possible factors, limitations, and pedagogical implications for using scenery-based VR in dyadic learning for ETP are discussed. Finally, the study suggests future research directions to combine technology-enhanced language learning theories with multimodal input and material design supported by immersive VR to enhance language production for specific purposes.