BACKGROUND: Dementia is associated with poor quality of life, but generic measures do not fully capture main aspects of these conditions. Our aim was to validate a Spanish version of the "dementia-specific health-related quality of life" DEMQOL system. METHODS: We validated the Spanish DEMQOL and DEMQOL-Proxy in 119 people with dementia from six centers and their caregivers. Patient-reported information, including generic (WHOQOL-BREF) and dementia-specific (DEMQOL) health-related quality of life, depressive symptoms (GDS-15), functional ability (Barthel Index), subjective perception of health and demographic information, was obtained by interview. Proxy-reported information from a family caregiver about the patient's quality of life (DEMQOL-Proxy) was also obtained by interview. RESULTS: The Spanish DEMQOL showed good psychometric properties. Acceptability (14.3% missing data), internal consistency (Cronbach's alpha = 0.85) and test-retest reliability (ICC = 0.71) were good. A priori hypotheses about the relationship between DEMQOL and the WHOQOL-BREF (psychological and physical domains), GDS-15 and Barthel Index were confirmed, indicating good construct validity. Associations between DEMQOL-Proxy and other measures were weaker. A three-factor solution accounted for 44.4% of the total variance of DEMQOL, and a four-factor solution accounted for 53.3% of the total variance of DEMQOL- Proxy. CONCLUSION: The Spanish versions of DEMQOL and DEMQOL-Proxy are reliable and valid in patients with mild/moderate dementia who are living at home.