We document substantial gender homophily in economics coauthorship and show how, in a male-dominated field, it generates unequal access to elite collaborators. Using data on over 2,800 faculty in the top 100 U.S. economics departments from 2010-2023, we find that female economists collaborate with men in 70 percent of their coauthorship pairs, 10 percentage points below a random-matching benchmark, a pattern that persists after accounting for field, cohort, and training. Homophily is amplified in high-stakes settings: men exhibit stronger same-gender sorting in top-five journal publications. These patterns translate into gender gaps in access to elite collaborators (top-decile publishers or faculty in top-ranked departments). Because the number of elite women is small, women's same-gender elite network is thin and highly concentrated: removing the ten most connected elite women cuts women's access to elite female coauthors by nearly two-thirds. In contrast, men's elite same-gender network is deep and diversified. A simple model demonstrates how symmetric same-gender preferences can generate asymmetric outcomes in a field with a male-dominated elite, without requiring discrimination. Using the empirical relationship between elite coauthorship and top-five publications, a bounding exercise implies that differential access to elite collaborators can account for up to 28 percent of the gender gap in top-five publications.