In recent years, many psychologists have become increasingly concerned about issues of reproducibility and replicability. From small sample sizes to post-hoc analytic flexibility ("p-hacking"), many factors conspire to decrease the robustness and trustworthiness of results in published research. These problems are important in developmental psychology as well, though the scope of the problem is unknown. I'll present some scientific and meta-scientific work I've done on these issues (including progress on the ManyBabies project, a large collaborative replication project in the infancy field) and describe some practical steps to take for increasing the robustness of your own work.