A longitudinal study of the functioning of top management teams in 27 hospitals examined relationships between group and organizational factors and team innovation. A model of group inputs, processes, and outputs was used, and it was predicted that group size, resources, team tenure, group processes, and proportion of innovative team members would affect the level and quality of team innovation. The results suggested that group processes best predict the overall level of team innovation, whereas the proportion of innovative team members predicts the rated radicalness of innovations introduced. Resources available to teams do not predict overall team innovation. The quality of team innovation (radicalness, magnitude, and novelty) may be determined primarily by the composition of the team, but overall level of innovation may be more a consequence of the team's characteristic social processes.