Objectives: To assess the occurrence and risk indicators of depression, anxiety, and comorbid anxiety and depression among nursing home patients and to determine whether depression and anxiety are best described in a dimensional or in a categorical classification system. Methods: DSM and subthreshold anxiety disorders, anxiety symptoms, major and minor depression and depressive symptoms were assessed in 333 nursing home patients of somatic wards of 14 nursing homes in the north west of the Netherlands with the Schedules for Clinical Assessment in Neuropsychiatry (SCAN) and the Geriatric Depression Scale (GDS). Comorbidity was studied along a severity gradient. Logistic regression analyses were carried out to identify demographic, health-related, psychosocial and care-related correlates of anxiety and depression. Results: The prevalence of pure depression (PD) was 17.1%, of pure anxiety (PA) 4.8%, and of comorbid anxiety and depression (CAD) 5.1%. Comorbidity increased dependent on severity of both anxiety and depression. Different patterns of risk indicators were demonstrated for PA, PD and CAD for the investigated baseline characteristics. Conclusions: Comorbidity of anxiety and depression is most prevalent in the more severe depressive and anxious nursing home patients. The gradual increase of comorbidity of anxiety and depression dependent on the levels of severity of depression and anxiety suggests that for nursing home patients a dimensional classification of depression and anxiety is more appropriate than a categorical one. The observed differences in patterns of risk indicators for PA, PD and CAD support a distinguishing of anxiety and depression. Future studies are needed to assess the effect of treatment of PA, PD and CAD in nursing home patients. Copyright © 2005 John Wiley & Sons, Ltd.