Questions about the reproducibility, replicability, and generalizability of the scientific literature have created anxiety within the scientific community and in the lay public (e.g., Open Science Collaboration, 2015). The ManyBabies project (Frank et al., 2017) is a productive response to these concerns: A global community of labs is actively collaborating to replicate and extend important research findings in developmental psychology, with a parallel goal of implementing and encouraging best practices (e.g., Open Science, pre-registration, no unethical data-peeking). Here we report on the current state and future directions of ManyBabies projects. The talks represent diverse theoretical perspectives, but stem from a unified collaborative framework that promotes robustness in research on early child development. The first talk reports results and lessons learned from the first ManyBabies study, ManyBabies1, which examined infants' preference for infant-directed speech (IDS). As the first large-scale replication attempt in developmental psychology, this study revealed key insights that future large-scale collaborations could build on. Second, a spin-off study explores the influence of bilingual language experience on the IDS preference. The third talk reports on ManyBabies2, which tests Theory of Mind in early childhood to contribute to recent controversial theoretical and methodological debates in the field. The final talk introduces ManyBabies4, which examines infants' preference for prosocial agents using a highly controlled design in order to address recent replication failures in a systematic way. Together, the talks will encourage further community-building by reporting on a successful project, updating about ongoing projects, and inviting participation in planned projects.