BACKGROUND: Anxiety and depression occur frequently, and recognition of their symptoms can be difficult because of comorbid medical conditions. The purpose of this study was to analyze the relations among symptoms, diagnoses, and severity of illness as indicators of anxiety and depression in primary care. METHODS: This was an observational, cross-sectional study of adult patients in a large, academic family medicine clinic. Patients completed the Duke Health Profile, which measures health-related quality of life and screens for anxiety and depression. Providers recorded patient diagnoses and assessed severity of illness. RESULTS: Patients with higher levels of anxiety and depression symptoms were more likely to have the diagnoses of headache, osteoarthritis, abdominal pain, and diabetes mellitus. These diagnoses, however, were no longer highly associated with anxiety and depression after controlling for age, sex, payer status, perceived health, pain, and disability. The indicators of high anxiety and depression symptom levels that persisted after controlling for all the other variables were female sex, low perceived health, more pain, and greater disability. CONCLUSION: In a primary care setting, female sex, self-reported perceived health, pain, and disability were more predictive of anxiety and depression than any of the most prevalent medical illnesses. Primary care providers need to be knowledgeable about these health measures so they can recognize patients at risk for anxiety and depression regardless of their medical diagnoses.