This study investigates how input modes and text types influence Korean college learners’ CALL listening and their perception of their concentration, comprehension and anxiety. The current study employed three input modes (listening-only, listening-with-subtitles, and listening-with-transcripts) and two text types (lectures and dialogues). The major findings are as follows: 1) subtitles or a transcript enhanced learner’s listening comprehension; 2) subtitles were the most helpful input mode whereas transcripts may have imposed more cognitive load to decode the reading; 3) the text types influenced CALL listening comprehension, but the familiarity effect may have been more dominant for more literate text types such as lectures; 4) apart from the higher English proficiency student’s perceived concentration, the two written input modes were perceived as beneficial tools for their concentration, comprehension, and reduced anxiety; 5) in particular, the lower English proficiency students tended to perceive the transcript as advantageous although they did not actually benefit from the transcript for their listening comprehension as much as they perceived. The pedagogical implications of the present study is subtitles and transcripts can be efficient pedagogical tools to draw learners’ attention to linguistic features or organization of the content to foster language ability. When designing CALL listening classes or applications, individual learners’ differences should be taken into account as well.