This quasi-experimental quantitative study examined the effects of different multimedia learning environments on the learning outcomes of second language learners. The study was designed to specifically assess Mayer's redundancy principle by investigating whether there was a difference in performance outcomes of second language learners participating in two different multimedia learning environments: 1) images with audio only and 2) images with audio and on-screen text. A repeated-measures analysis of covariance (ANCOVA) was conducted to evaluate change in comprehension performance from pretest to posttest, mean differences in comprehension performance between the two groups, and group differences in comprehension growth from pretest to posttest, using gender and English proficiency as control variables. Additionally, an analysis of covariance (ANCOVA) was also carried out to assess group differences in posttest scores, controlling for pretest scores, gender, and English proficiency. Results showed no statistically significant difference in knowledge growth between participants in the two conditions and no statistically significant difference between groups in knowledge posttest scores. However, there was a significant change from pretest to posttest for the combined sample of the learners. Additionally, a significant group × gender interaction effect was observed, where males performed better on the knowledge posttest under the Images with Audio condition while females performed better under the Images with Audio and On-Screen Text condition. Results from this study suggested that for second language learners, the presence of on-screen text, although redundant to the information contained in the audio, did not appear to impair their learning.