Intense competition for limited opportunities means the career path of a scientist is a challenging one, and female scientists in particular are less likely to survive in academia. Collaboration is a key factor in scientific advances, and in social species enhanced sociality improves fitness and longevity. Yet whether sociality influences career progression and survival in science, and how this might differ between genders, is largely unknown. We built authorship social networks from publication records to test how sociality predicts career progression and survival in a cohort of biologists contributing to three international conferences in the 1990s. We show that sociality has the strongest effect for female researchers but, regardless of gender, publishing with many diverse co-authors significantly reduces time to become a principal investigator and increases career duration. Publishing repeatedly with co-authors also enhances career progression in both genders, but reduces career length for men. These findings demonstrate that the value of collaboration extends beyond scientific advances, and can directly benefit the career progression and longevity of research scientists themselves. Efforts to encourage researchers at all levels to invest in collaborations, particularly with female researchers, will help to close the gender gap in science and academia.