Drug repositioning or repurposing has received much coverage in the scientific literature in recent years and has been responsible for the generation of both new intellectual property and investigational new drug submissions. The literature indicates a significant trend toward the use of computational- or informatics-based methods for generating initial repositioning hypotheses, followed by focused assessment of biological activity in phenotypic assays. Another viable method for drug repositioning is in vitro screening of known drugs or drug-like molecules, initially in disease-relevant phenotypic assays, to identify and validate candidates for repositioning. This approach can use large compound libraries or can focus on subsets of known drugs or drug-like molecules. In this short review, we focus on ways to generate and validate repositioning candidates in disease-related in vitro and phenotypic assays, and we discuss specific examples of this approach as applied to a variety of disease areas. We propose that in vitro screens offer several advantages over biochemical or in vivo methods as a starting point for drug repositioning.