Despite advances in early detection and treatment, colon cancer remains a leading cause of cancer death in the United States. Traditional clinical and pathological features of colon cancer are inadequate in predicting survival. Patient characteristics, such as germline genetic variation, may provide additional prognostic information. To date, only a few studies have used candidate gene and pathway approaches to identify germline genetic factors related to cancer outcomes, and no comprehensive genome-wide studies have been conducted to assess broader genetic variation in relation to outcomes after colon cancer diagnosis. Randomized clinical trials (RCTs) provide a powerful setting in which to address this gap in knowledge, because the patient populations are well-characterized, treatment is standardized, and follow-up is uniform. The overarching goal of the proposed study is to identify genetic factors associated with colon cancer clinical outcomes and to assess whether incorporation of genetic variants improves existing prognostic models. To achieve this goal, we will leverage the resources of three NCI-sponsored phase III RCTs of colon cancer and a recently approved collaboration with the Center for Inherited Disease Research. We propose to use a discovery-based search of host genome-wide single nucleotide polymorphism data in over 6,500 stage II-III colon cancer patients participating in three RCTs, all of whom received 5-fluorouracil / leucovorin / oxaliplatin (FOLFOX) chemotherapy with or without adjuvant therapy, to examine associations with clinical outcomes. Specifically, we will evaluate associations between common genetic variation and disease-free and overall survival among patients with stage II-III colon cancer (Aim 1). We will also evaluate germline genetic variation in relation to treatment-associated serious adverse events (Aim 2). Finally, we will evaluate the impact of adding genetic data into existing web-based, publicly available prognostic models to determine if germline genetic loci can be used, in combination with patient characteristics and clinical factors, to more accurately predict colon cancer outcomes (Aim 3). These prognostic models will be validated in an independent sample of 1,000 stage II-III colon cancer patients from the three RCTs. Data yielded by these investigations will be among the first describing genome-wide germline genetic factors associated with colon cancer prognosis. These results have high translational potential for informing prognosis, augmenting both current strategies that rely on traditional clinical factors and emerging strategies that incorporate information on somatic molecular alterations. Moreover, identifying loci associated with clinical outcomes after colon cancer diagnosis may provide mechanistic insight into cancer progression and metastasis, potentially illuminating new therapeutic targets that can be exploited for clinical translation.