Self-report is the dominant measurement approach in psychology, but self-report items are often constructed in ways that limit their usefulness. We review basic principles of measurement, which we then develop as a framework to direct the construction of self-report items. This framework calls on researchers to define the nature of the construct of interest (attending to its hypothesized breadth and dimensionality), to articulate the intended item-operating characteristics of items administered, and to implement strategies to reduce the influence of both systematic and random sources of error. We provide guidance on the writing of questions, the construction of response metrics, and the aggregation of data from multiple items to produce scale totals. Attention to the issues reviewed in this chapter will help researchers write self-report items that will help them answer their own questions.