Introduction: Women have historically been under-represented in medical literature, particularly prominent in authorship of invited commentaries. With the instantaneous change in work environment forcing Americans to adapt to working at home, many theorize that women will be more adversely affected due to traditional concepts of women being more responsible for the home in addition to work responsibilities. Objective: Understand how women contributed to coronavirus disease 2019 (COVID-19) literature early in the pandemic. Methods: Cross-sectional analysis of manuscripts published in three high-impact U.S. medical journals from February to May 2020 was performed. We used log-binomial regression to examine for an association between COVID-19 status and likelihood of having at least one female first author, and evaluated for effect modification according to whether the manuscript was invited. Results: Among 980 manuscripts, 313 (31.9%) listed at least one female first author, 203 were written on COVID-19 (20.7%), and 144 (14.7%) were invited. There was no association between COVID-19 status and having at least one female first author overall (adjusted risk ratio [RR] 0.93, 95% confidence interval [CI] 0.72-1.19). The relationship between COVID-19 status and first-author sex was 0.23 (95% CI 0.06-0.92) for invited manuscripts and 1.04 (95% CI 0.81-1.35) for noninvited manuscripts (p for interaction 0.02). Discussion: We demonstrate that women were not less likely to be first authors on COVID-19 manuscripts but were less likely to be first authors on invited COVID-19 manuscripts. Early career female researchers are the most vulnerable for inability to meet metrics for promotion, accounting for the continued under-representation of women in senior academic roles. COVID-19 has the potential to exacerbate this disparity.