This study investigates whether a modified version of the COOP/WONCA charts is suitable to assess quality of life (QOL) in persons with dementia in nursing homes. A group of 112 institutionalized persons with moderate to severe dementia were approached for an interview. Twenty-two were observed not to be communicative, leaving 90. Sixty-seven persons were able to answer four out of six questions adequately (interviewable). Inter-observer reliability (n=38) was excellent (weighted kappa 0.90 to 0.97). Test-retest reliability (n=34; one week interval) ranged from poor for Daily and Social Activities and for the QOL charts, to moderate for Feelings and Pain, and satisfactory for Physical Functioning (weighted kappa 0.23 to 0.67). Interviewability was associated with severity of the dementia and communication ability. Support for convergent validity was found in medium-sized Spearman correlations between the COOP/WONCA charts and related variables. Support for discriminant validity was found in the absence of association between the Ccharts and non-related variables. The modified COOP/WONCA charts can be used to assess QOL in 60% of people with dementia in nursing homes but further modification is needed. Severe cognitive impairment and communication disabilities proved limiting factors for the use of the instrument. Although the illustrations on the charts appeared not to be helpful, the written response options in addition to verbal presentation proved useful during the administration of the charts.