This paper undertakes a conceptual and methodological examination of construct validation from the viewpoint of a scientific realist philosophy of science. It is argued that there is an important need to revise our thinking about the methodology of construct validation and to incorporate the revisions into clinical psychological practice. After a brief characterisation of construct validation, the paper points out the inadequacies of operationalism and cautions against the heavy appeal to definitions in science. Thereafter, psychology's standard construct validation strategies of hypothetico-deductive testing and null hypothesis significance testing are criticised, and the importance of abductive reasoning for construct validation is spelled out.