Gender effects in large-scale assessments have become an increasingly important research area within and across countries. Yet few studies have linked differences in assessment results of male and female students in higher education to construct-relevant features of the target construct. This paper examines gender effects on students' economic content knowledge with a focus on construct-relevant explanations. Moreover, we compare gender effects cross-nationally between Germany, Japan, and the United States. To assess economic content knowledge of higher education students, we used translated, adapted, and validated versions of the Test of Understanding in College Economics (TUCE, 4th ed.), an instrument that is commonly used internationally. We found gender effects on test scores in all three countries; effects were larger in Germany and the United States than in Japan. Gender effects were generally more pronounced on the numeracy subscale than on the literacy subscale, that is, male students had a greater edge over female students when items required calculations. In our conclusion, we discuss how numeracy and literacy items may tap different abilities.