6505 Background: Lower SES is known to be associated with worsened cancer survival. Here we evaluate the impact of SES on stage of cancer at diagnosis in Ontario which has universal health insurance. Methods: All incident cases of breast, colon, rectal, non-small cell lung, cervical and larynx cancer diagnosed in Ontario 2003–2005 were identified using the Ontario Cancer Registry. Stage information is only captured routinely for patients seen at Ontario's 8 Regional Cancer Centers (RCCs). This represents approximately 68% of the population and forms the basis for all analyses. Using a best stage grouping approach, cases were assigned stage based on pathologic TNM if available and clinical TNM otherwise. The population of Ontario was divided into quintiles based on community median household income reported in the 2001 Canadian census. Using postal code at time of diagnosis cases were assigned to quintiles (Q); Q1 represents the communities where the poorest 20% of the Ontario population resided. Comparisons between Q1 and Q2–5 were made using the chi-square test. A Cox model was used to evaluate overall survival, SES, stage, and age. Results: Stage at diagnosis was available for 19,239/23,254 (83%) of cases seen at RCCs. Among cases with breast cancer, those in Q1 were less likely to have stage I disease (43 vs 47%, p = 0.004) and more likely to have stage IV disease (5 vs 4%, 0.008) than Q2–5. With lung cancer, cases in Q1 were more likely to have stage I disease compared to Q2–5 (16 vs 13%, p = 0.015). Distribution of stage I and stage IV disease did not differ by SES across other individual diseases. However, for all 6 cancers combined, cases in Q1 were less likely than Q2–5 to have stage I disease (27 vs 30%, p = 0.001) and more likely to have stage IV disease (21 vs 18%, p < 0.0001). We found significant gradients in 3-year overall survival across Q1-Q5 for breast (5% absolute difference in survival, p < 0.001), colon (4%, p = 0.049), and cervical (18%, p = 0.031) cancers. Adjustment for stage and age only slightly diminished these survival gradients. Conclusions: Despite universal health care, SES remains associated with survival among patients with cancer in Ontario. These data suggest that the difference in outcome is only partially explained by differences in stage at diagnosis. No significant financial relationships to disclose.