OBJECTIVE: This paper reports the impact of oral health on the quality of life (QOL) of Southern European, dentate older adults, living independently in Melbourne, Australia. Participants were recruited through ethnic social clubs and interviewed about oral health, general health, socio-demographics, and QOL using the Medical Outcomes Study Short-Form 12 (SF-12). The SF-12's physical and mental health component summary scores (PCS and MCS, respectively) were computed. The Oral Health Impact Profile (OHIP-14) assessed the specific impact of oral health on QOL. Participants were also given a clinical oral examination. RESULTS: A total of 603 eligible older adults volunteered; 308 were from Greek background and 295 were from Italian background. Mean age was 67.7 years (SD 6.2), with 63.7% being female. The PCS score had a mean value of 45.8 (SD 11.8), and MCS had a mean of 47.8 (SD 5.7). PCS was associated with, periodontal status, chronic health condition, self-perceived oral health needs, self-assessed oral health status, oral health impact score and the interaction between gender and level of education [F(11 552) = 10.57; P < 0.0001]. These independent variables accounted for 16% of the variance in PCS. The multivariate model predicting MCS had only one significant variable (self-reported gingival bleeding), explaining 1.5% of the variance. The OHIP-14 ranged from 0 to 48 with a mean score of 5.6 (SD 9.3). The model predicting OHIP-14 contained four significant variables: perceived oral health treatment needs, number of missing natural teeth, reports of having to sip liquid to help swallow food, and gender [F(4576) = 33.39; P < 0.0001], and explained 18% of the variance. The results demonstrated a negative association between oral health indicators and both the oral health-related QOL and the physical component of the SF-12. CONCLUSION: The present findings support a growing recognition of the importance of oral health as a mediator of QOL. However, the self-selected sample and modest predictive power of the multivariate models suggest that further research is needed to expand this explanatory model.