585 Background: Randomized clinical trials for second line treatment (Tx2) of metastatic colon cancer (mCC) often have strict inclusion/exclusion criteria regarding prior treatment, yet in the real world there is significant variation. This study aims to determine whether cost effectiveness estimates of Tx2 for elderly mCC patients varies by the regimen they received in first-line treatment (Tx1). Methods: We identified 3,211 elderly (age 66+) mCC patients in the SEER-Medicare dataset who received NCCN recommended Tx1 between 2003 and 2009. Patients were categorized by Tx1 based on a previously published algorithm as fluorouracil and leucovorin (5-FU/LV), irinotecan (IRI), oxaliplatin (OX), or “other,” which included IROX or biologics without OX or IRI. Separate 5-year incremental cost-effectiveness of Tx2 were calculated for each Tx1. Approximately 1% of patients with outlier costs were excluded. Patients enrolled in HMOs, lost Part A and/or B, and died of causes other than colon cancer are censored. We adjusted for censoring using the Inverse Probability Weighting (IPW) method. Costs were inflation-adjusted to 2009 dollars using the national monthly medical price index. Results: Among patients who received Tx1, 34% (n=1,090) received 5FU, 17% (n=530) received IRI, 46% (n=1,481) received OX, and 3% (n=110) received other (IROX or Biologics) regimens; 44.5% (n=1,440) proceeded to Tx2. Compared to those who do not receive Tx2, patients who received Tx2 following IROX or Biologics, IRI and 5FU in Tx1 live 292 (se = 4), 224 (se = 2), and 191 (se = 2) days longer and incur added costs of $49,096 (se = $7,137), $83,784 (se = $12,322), and $91,686 (se = $10,312), respectively. Recipients of OX in Tx1 did not receive a survival benefit from Tx2, despite additional costs of $46,849 (se = $10,468). Conclusions: The real-world survival benefit of Tx2 for elderly mCC patients in SEER-Medicare varied based on Tx1 from potential harm to a mean of 292 days of incremental survival. Similarly, the costs and cost effectiveness of Tx2 varied by Tx1. These results underscore the importance of considering prior treatment when evaluating the benefit of subsequent treatment for elderly mCC patients.