Qualitative methods are typically and formally used only in the earliest phases of instrument development to generate items. Once these items are generated, instrument development usually then moves into the testing phases, where quantitative methods prevail. The achievement of psychometric credibility is presumed to depend largely on quantitative measures of reliability and validity. Or if qualitative methods are employed, their use is masked, unfocused, and/or unplanned. The planned use of qualitative methods is critical in every phase of instrument use and in all studies that depend for their results on instruments, and their use is critical in illuminating problems with existing instruments. The authors illustrate these points by drawing on the first author's experiences in the field with the Beck Depression Inventory in her research program on managing fatigue in persons with HIV/AIDS.