BACKGROUND: The End-of-Life in Dementia (EOLD) scales comprise the most specific set of instruments developed for evaluations of patients' end of life by their families. It is not known whether the EOLD scales are useful for cross-national comparisons. METHODS: We used a mortality follow-back design in multi-center studies in the Netherlands (pilot study 2005-2007) and the U.S.A. (1999), and we compared EOLD Satisfaction With Care (SWC; last three months of life), Symptom Management (SM; last three months) and Comfort Assessment in Dying (CAD) scores for 54 Dutch and 76 U.S. nursing home residents. RESULTS: SWC total scores did not differ significantly between the Dutch and U.S. studies (31.9, SD 4.7 versus 30.4, SD 6.1), but three of ten items were rated more favorable for Dutch residents, as were SM total scores (29.1, SD 9.2 versus 20.4, SD 10.6). CAD total scores did not differ (32.0, SD 5.4 versus 30.5, SD 5.9, respectively), but the "well-being" subscale was rated more favorably for Dutch residents. Results were similar after adjustment for demographics and dementia severity. CONCLUSION: The Dutch families rated end of life with dementia in nursing homes as somewhat better than did U.S. families. Although differences were small, the observed patterns were consistent. This suggests validity of the SM and CAD to assess differences in quality of dying and possible sensitivity to differences between countries or time frames. Larger, simultaneous, cross-national studies are needed to confirm usefulness of the scales and to detect areas which need improvement in the respective countries