Four labs discovered and replicated 16 novel findings with practices such as preregistration, large sample sizes and replication fidelity. Their findings suggest that with best practices, high replicability is achievable.