At the end of phase 2 of drug development, a decision to initiate phase 3 studies is made that involves the commitment of considerable resource. One of the main reasons of Phase 3 failure in drug development is overoptimistic assumptions for the planning of phase 3 resulting from limited phase 2 information and/or unawareness of realistic success probabilities. In this round table discussion, we like to share experiences, questions, and models for go/no go decision criteria setting that go beyond simplistic point estimates or the traditional hypothesis test approach of type I and type II error rates.