Up to now, only a few studies in multimedia learning have focused on gender effects. While research has mostly focused on learning success, the effect of gender on instructional efficiency (IE) has not yet been considered. Consequently, we used a quasi-experimental design to examine possible gender differences in the learning success, mental effort, and IE of college preparatory school students (N = 218, German 9th graders; 93 young men, 125 young women). We established three groups: a treatment group (computer-based learning unit, CBLU); a control group (textbook-based version, TBLU); and an external control group (no instruction). The subject taught was genetically produced insulin. We applied a pretest-posttest-retention test design. We used the long-term cognitive achievement and the mental effort invested during the retention test phase to examine the IE. Young women and young men differed in IE with a small effect size, with the young women achieving significantly higher long-term learning success. Our results indicate that the applied instructional design of the CBLU seems particularly beneficial for young women's cognitive achievement. (Contains 7 tables, 2 figures, and 1 footnote.)