To the Editor: We appreciate the letter by Dr. Kales and her colleagues commenting on our recent report. They focus on an important factor that should be addressed by both clinicians and clinical investigators—namely, the documented differences in the recognition of depression by race in treatment settings. The studies they note, however, refer exclusively to treatment samples. Three of the samples (Fabrega et al., 1994, Mulsant et al., 1993, Leo et al., 1997) were derived from psychiatric inpatient populations. The use of treatment samples to generalize to the community (from which our sample was drawn) can potentially lead to bias. For example, older African Americans may be admitted differentially to psychiatric facilities because of differences in their seeking of health care, which limits the ability to conclude that these African American elders are less likely to receive a diagnosis of major depression in a clinical setting. African Americans may be just as likely to receive a diagnosis of major depression in an ambulatory setting, but they may be less likely to be hospitalized (thus leading to the higher rates of admission for dementia and schizophrenia reported). Unfortunately, our data and the data from the studies cited by Dr. Kales and colleagues cannot address these questions.