This chapter discusses adaptive two-stage dose-finding designs. Two-stage designs can be developed to address the objectives sequentially, e.g., including a proof of concept (PoC) objective within a dose-finding study. The chapter introduces statistical dose–response model, and defines the Fisher information for a two-stage design. It discusses six different two-stage designs using the chronic obstructive pulmonary disease (COPD) study as an illustration. Objectives of Phase II dose-finding studies include the proof of drug-related effects, estimation of the dose–response, and the selection of doses for confirmatory testing. Interim timing defines the number of patients allocated to the first and second part of the two-stage design. Dette et al. study the efficiency of two-stage designs depending on the interim timing using asymptotic considerations and analytical approximations to the interim uncertainty. Their idea of second-order Taylor approximations to the information matrix might also be extended to design criteria including model uncertainty, as in MCP-Mod designs.