This quasi-experimental quantitative study investigated the effects of different formats of multimediaenhanced instruction on knowledge gain and retention of second language learners. The study focused specifically on assessing Mayer's redundancy principle. Second language learners were randomly assigned into two groups and were presented with two formats of a multimedia-enhanced lesson: (1) images with audio and (2) images with audio and on-screen text. A repeatedmeasures analysis of covariance (ANCOVA) was carried out to evaluate change between pretest and immediate posttest, mean differences in knowledge gain between the two groups, group differences in growth from pretest to immediate posttest, change between immediate posttest and delayed posttest, and group differences in knowledge retention as indicated by difference scores between immediate posttest and delayed posttest, using gender and English proficiency as control variables. Additionally, an independent samples t-test was performed to assess the effect of the two formats of multimedia-enhanced instruction on the learners' knowledge retention. Results showed that participants engaging with both formats of multimedia-enhanced instruction improved significantly from pretest to posttest. Moreover, participants who engaged with the multimedia-enhanced instruction where on-screen text was included were also found to perform significantly better than participants who were in the group where on-screen text was not present. Second language learners in both groups were also able to retain their knowledge at the same level. These results appear to contradict the redundancy principle as having onscreen text, which is considered redundant, presented together with images and audio in a multimedia-enhanced lesson was not detrimental to learning.