Incidental vocabulary learning is often seen as superior to direct instruction on many occasions. Meanwhile, upon the emergence of the World Wide Web, second language (SL) learners have been introduced to 'podcasts' (recorded audio and video online broadcasts) which could be authentic sources of vocabulary learning. The relatively recent phenomenon of video podcast (vodcast) might be considered as a reliable complementary source of input to the written text or the audio track which are predominantly used to represent the platforms of SL instruction. To examine this assertion, three groups of Iranian EFL learners (n=63) were independently exposed to different modes of input (the reading text, audio track, and vodcast) during a series of classroom sessions under highly controlled circumstances. Immediate and delayed passive recall tests of vocabulary were administered to investigate their incidental gains. A multivariate analysis of variance revealed that both dependent variables (immediate and delayed recall) were significantly affected by the input modes. The post-hoc tests indicated no significant difference between the written and the audio groups while the vodcast group significantly outperformed the other two. The rich contextual clues made available by this audiovisual source seem to account for its superiority.