We implemented instruction on the role of metacognition and on metacognitive strategies in an undergraduate, upper-division, two semester Human Anatomy and Physiology sequence. We previously reported that metacognitive training improved self-reported metacognitive performance and buffered decreases in self-efficacy but did not improve measures of academic performance for these students (Schaefer, 2018). This work expands on those findings to examine whether metacognitive training differentially impacted male vs. female student metacognitive practices, self-efficacy, or academic performance. Metacognitive ability and self-efficacy are well-documented to have direct relationships with academic success (Bandura, 1977; Flavell, 1979; Nickerson et al, 1985; Multon et al., 1991; Pajares, 1996; Georghiades, 2000; Zimmerman, 2000; Flavell et al., 2002; Tanner, 2012; Trujillo and Tanner, 2014). Self-efficacy and metacognitive ability have also been linked to each other, but whether they have a causative relationship or are linked via an indirect mechanism is unknown (Houtveen et al., 2004; Cera et al., 2013). If metacognition and self-efficacy have a synergistic effect on academic performance, the effect would be especially important in courses perceived as “challenging” (such as upper-division biology courses) because self-efficacy is situation-specific such that general academic ability may not determine self-efficacy in the challenging course. Male and female students entered the first semester of the course with no difference in self-reported metacognitive performance between genders, measured by the Motivated Strategies for Learning Questionnaire Metacognitive Self-Regulation scale (Pintrich et al., 1991). Female students entered the course with higher GPAs while male students entered with higher self-efficacy. Students were split into a metacognitive instruction group and a control group based on course section enrollment. The metacognitive treatment group received regular instruction in metacognitive strategies and in the role of metacognition for academic success. The control group received no explicit instruction on metacognition. Students of both genders who received metacognitive instruction improved self-reported metacognitive practices while students who did not receive metacognitive instruction did not improve. Therefore, metacognitive instruction was effective for both genders. Students of both genders reported decreased self-efficacy, measured by the New General Self-Efficacy scale (Chen et al., 2001), and increasingly correlated self-efficacy with exam scores throughout the semester. Male and female final grades and average exam scores strongly correlated with end of course self-efficacy in the fall semester. Female students, but not male students, demonstrated a correlation between average exam scores and beginning of course self-efficacy. Metacognitive instruction did not improve exam scores or final course grades in either gender.