The objectives of this study were to investigate: (a) multiple relations between socio-demographic, psychosocial, and health variables and quality of life in the elderly and (b) the model's validity through correlation with depressive symptoms. The sample included 339 elderly individuals from 60 to 98 years of age (M = 73.4; SD = 8.3), who answered a socio-demographic questionnaire, WHOQOL-100, and BDI. Cluster analysis of the sample distributed the elders into two groups according to self-perceived quality of life (better versus worse), and logistic regression analysis identified variables that explained better quality of life. Social class, self-rated health status, volunteer work, use of medication, and data collection setting were associated with quality of life (predictive capacity for correct classification 72.3%, specificity 73.6%, and sensitivity 71.1%). The inverse correlation between the model's variables and BDI scores provided evidence of the model's validity. The model can help support public policies aimed at promoting quality of life in the elderly.