Background: Little is known about the quality of life of people with very severe dementia in long-term care settings, and more information is needed about the properties of quality of life measures aimed at this group. In this study we explored the profiles of quality of life generated through proxy ratings by care staff and family members using the Quality of Life in Late-stage Dementia (QUALID) scale, examined factors associated with these ratings, and further investigated the psychometric properties of the QUALID. Methods: Proxy ratings of quality of life using the QUALID were obtained for 105 residents with very severe dementia, categorised as meeting criteria for Functional Assessment Staging (FAST) stages 6 or 7, from members of care staff (n = 105) and family members (n = 73). A range of resident and staff factors were also assessed. Results: Care staff and family member ratings were similar but were associated with different factors. Care staff ratings were significantly predicted by resident mood and awareness/responsiveness. Family member ratings were significantly predicted by use of antipsychotic medication. Factor analysis of QUALID scores suggested a two-factor solution for both care staff ratings and family member ratings. Conclusions: The findings offer novel evidence about predictors of care staff proxy ratings of quality of life and demonstrate that commonly-assessed resident variables explain little of the variability in family members? proxy ratings. The findings provide further information about the psychometric properties of the QUALID, and support the applicability of the QUALID as a means of examining quality of life in very severe dementia.