This chapter introduces methods for evaluation of automatic question generation. There are two reasons for presenting evaluation before we discuss AQG techniques. First, we will mention evaluation results when we present the various AQG methods in subsequent chapters, so familiarity with evaluation methods will be needed. Second, evaluation methods involve not only strict methodology; they highlight the more general approaches toward what is important, what are the valued outcomes in the generation process. It may be useful to become acquainted with this variety before we delve into the specific generation methods. In this chapter we describe common human-based intrinsic evaluation methods, common intrinsic automated evaluation methods, and some approaches to extrinsic evaluation of AQG.