The objective of this study was to investigate the relationship between continuity of family physician (FP) care and inpatient hospitalizations in elderly people with diabetes who have universally-insured health care. We constructed a population-based retrospective cohort study using a sample of 1143 people aged 65 years or older with newly diagnosed diabetes who were selected from a longitudinal surveillance database in the province of Newfoundland and Labrador (NL), Canada. Continuity of FP care was estimated by 3 chronological indices (Continuity of Care [COC], Usual Provider Continuity [UPC], and Sequential Continuity [SECON]) using administrative physician claims data. Age, sex, number of chronic conditions, and income were used as control variables. People with high continuity had lower crude rates of hospitalization than those with lower continuity. Log-linear regression analysis showed that higher continuity was associated with decreased rates of hospitalization in an unadjusted model [rate ratio (95% confidence interval)]; COC: 0.73 (0.61-0.86); UPC: 0.71 (0.59-0.86); SECON: 0.64 (0.52-0.78), and after adjusting for control variables; COC: 0.82 (0.69-0.97); UPC: 0.82 (0.68-0.98); SECON: 0.75 (0.61-0.91). Other significant predictors of reduced hospitalizations were female sex, fewer chronic conditions, and higher income. The findings suggest that high levels of continuity of FP care are associated with reduced hospitalizations in elderly people with diabetes within a universally-insured health care system.