Recent CALL research has started to go beyond the comparison of CALL versus non-CALL environments to explore what learners do while going through CALL activities. One important strand within that area has focused on the use and utility of providing learners with opportunities to request modified input. While many studies have carried out research in a laboratory setting or through structured observation, this study explores learners' interactions within a CALL program in an authentic setting. The results indicate large variation in the use of options available. While teacher-introduced and mandated options were utilized more frequently, the data revealed that some options were either used infrequently (e.g., access to a textual gloss), or completely ignored by half of the learners (i.e., simultaneous repetition of text and audio). However, the study also found that access to options that provide added redundancy were significant predictors of success, more so than the time spent interacting with the CALL program. While low proficiency students chose dual input (audio and textual), they appeared to be less able to effectively utilize it. Higher proficiency students mainly focused on audio repetition only. (Contains 1 figure and 9 tables.)