This study investigated learner behavior and performance in a CALL multimedia listening activity which offered a video and two textual help options - a lecture transcript and subtitles in cases of comprehension breakdowns. In particular, the study examined and compared the learners' use of two help options by looking at the time and frequency of interaction with help. The study aimed to identify and compare possible patterns of participants' interaction with help options as well as investigate possible differences between participants at two proficiency levels. The participants in the study were eighteen ESL college students at the intermediate level of listening proficiency who were enrolled in an academic listening class. The materials given to participants included two tests, two questionnaires, and the CALL listening comprehension activity entitled the Astronomy unit. Students' work on the Astronomy unit was recorded using Camtasia screen capturing program and students' interaction with the material later transcribed. The results showed that participants varied in their use of help options in terms of time on help, number of help page openings, and number of instances of useful interaction with help. Generally, the students interacted with the subtitles more frequently and for longer periods of time than with the transcript. Furthermore, the students exhibited four different patterns of interaction with help: subtitles, transcript, non-interaction, and mixed interaction pattern. While the subtitles and the transcript pattern groups showed very similar behavior, the non-interaction group differed from other groups the most. The differences between the two proficiency groups (higher and lower) were found on their performance during and after the activity with the higher group having better comprehension. The findings obtained showed that learners don't always take advantage of help features in CALL listening materials and that software design as well as teachers' use of software with students could promote interaction with help. The results suggested that in post-comprehension breakdowns, learners should be given both the subtitles and the transcript as well as a choice to skip help. Further research could look at a larger and varied learner sample in addition to the use of materials in different settings.