Given the accumulating evidence about audiovisual input as a valuable resource from which knowledge of multiword expressions (MWEs) can be built up incidentally, the next inquiry arises as to what can be done to promote MWE uptake from this resource. Despite the increasing popularity of bilingual subtitles as a form of on-screen text, their effectiveness for the incidental acquisition of MWEs relative to other subtitling forms has not yet been examined. A total of 89 L2 learners were randomly assigned to three experimental groups that differed in terms of the subtitling condition (captions, bilingual subtitles, L1 subtitles) under which they viewed an input video containing target MWEs twice. They were administered a pretest and a posttest to gauge their improvement in MWE knowledge at the level of form recognition and meaning recall. An operation-span task was employed to measure their working memory capacity. Results from generalized linear mixed-effects models revealed that bilingual subtitles had an advantage over captions and L1 subtitles in facilitating MWE meaning uptake, and they were as effective as captions in promoting MWE form uptake. Working memory played a predictive role in the uptake of novel MWEs, with a greater weight observed in bilingual subtitled viewing.