Collaborations among researchers and across disciplinary, organizational, and cultural boundaries are vital to address increasingly complex challenges and opportunities in science and society. In addition, unprecedented technological advances create new opportunities to capitalize on a broader range of expertise and information in scientific collaborations. Yet rapid increases in the demand for scientific collaborations have outpaced changes in the factors needed to support teams in science, such as institutional structures and policies, scientific culture, and funding opportunities. The Science of Team Science (SciTS) field arose with the goal of empirically addressing questions from funding agencies, administrators, and scientists regarding the value of team science (TS) and strategies for successfully leading, engaging in, facilitating, and supporting science teams. Closely related fields have rich histories studying teams, groups, organizations, and management and have built a body of evidence for effective teaming in contexts such as industry and the military. Yet few studies had focused on science teams. Unique contextual factors within the scientific enterprise create an imperative to study these teams in context, and provide opportunities to advance understanding of other complex forms of collaboration. This review summarizes the empirical findings from the SciTS literature, which center around five key themes: the value of TS, team composition and its influence on TS performance, formation of science teams, team processes central to effective team functioning, and institutional influences on TS. Cross-cutting issues are discussed in the context of new research opportunities to further advance SciTS evidence and better inform policies and practices for effective TS.