A substantial body of research shows that various types of audiovisual (AV) input such as videos and videos with second language (L2) subtitles can facilitate language learning. However, language teachers' day-to-day practices with regard to multimodal input is less understood. To bridge the gap in language education, this study investigates teachers' perceived use of four types of AV input (video only, video with subtitles, video with captions and video with enhanced captions) and factors influencing teachers' perspectives on these types of input for in-class and out-of-class learning. Questionnaire data were collected from 193 L2 teachers across the globe about their perceived use of AV input. Teachers reported that they use video and captioned video most frequently in both classroom and out-of-class contexts. Logistic regression analyses revealed that teachers' perceived importance and comfort using specific AV input types were the two most important factors explaining teachers' reported use. Complementarily, open-ended responses were analyzed qualitatively to identify teachers' additional reasons for (non)use of such input.