The assessment and management of pain in nursing homes have been shown to be suboptimal, but no study has evaluated differences in clinical setting within these homes. The prevalence and management of pain on different care wards (psychogeriatric, somatic, and rehabilitation) was studied on 562 newly admitted Dutch nursing home residents. Pain was measured according to the Nottingham Health Profile (perceived pain) and the Minimum Data Set pain observation items (frequency and intensity). Pain frequency differed significantly across the different ward types: on psychogeriatric wards (n = 247), it was 27.1%; on somatic wards (n = 181), 53.9%; and on rehabilitation wards (n = 129), 57.8%. Being admitted on a psychogeriatric ward was significantly related to less pain compared to being admitted on a somatic ward, even when adjusted for possible confounders such as age, gender, cognitive status, activities of daily living, pain-related disorders, and depression (odds ratio [OR] 0.38 [95% confidence interval (CI) = 0.23-0.62]). Patients on psychogeriatric wards who had pain received less pain medication, adjusted for frequency and intensity of pain (OR 0.37 [95% CI = 0.23-0.59]), compared to patients on somatic wards. We conclude that admission to a psychogeriatric care ward, independent of cognition, is associated with lower pain prevalence, and also with lower levels of pain treatment. © 2007 U.S. Cancer Pain Relief Committee.