Background: Approximately 14% of CRC patients harbor a germline pathogenic variant (gPV) in a cancer susceptibility gene. As gPV prevalence alone is insufficient evidence for cancer causation, we performed an integrated germline and somatic genomic analysis to discover the extent to which gPVs may be driving CRC carcinogenesis. Methods: Using the MSKCC IMPACT dataset, we identified all patients with CRC who had undergone targeted tumor-normal sequencing (NCT01775072) inclusive of up to 90-genes assessed for germline variants. gPV in tested genes were categorized according to known association with CRC susceptibility and penetrance level. For example, high/moderate penetrance CRC-risk genes included: APC, MLH1, MSH2, MSH6, PMS2, EPCAM, biallelic MUTYH, SMAD4, STK11, BMPR1A, PTEN, POLD1, POLE, TP53. Low-penetrance CRC-risk genes included: monoallelic MUTYH, APC I1307K, CHEK2, BLM. In patients with a gPV, biallelic inactivation at the implicated germline region was inferred based on loss of heterozygosity (LOH) or somatic pathogenic alteration in the tumor. Results: Of 3988 CRC patients, 16.3% (650/3988) harbored at least one gPV (65 patients: 2 gPVs; 3 patients: 3 gPVs). Among a total of 721 gPVs encompassing 55 genes (252 high-penetrance, 118 moderate-penetrance, 351 low/recessive/uncertain-penetrance), overall biallelic inactivation was observed in 41%. In patients with gPVs associated with high/moderate penetrance CRC-risk genes, biallelic inactivation in tumors was present in 70%, while in patients with gPVs in high/moderate penetrance non-CRC genes this was 24% (p = < 0.0001). When assessed according to age at CRC diagnosis, biallelic inactivation in high/moderate penetrance CRC-risk genes was higher in patients diagnosed at age < 35 than in patients age > 50 (p = 0.003); no such difference was observed in high/moderate penetrance non-CRC-risk genes. Among non-CRC-risk genes, in the high/moderate penetrance HRD genes, biallelic loss in the tumor was seen in 29% gATM, 54% gBRCA1, 22% gBRCA2, 30% gBRIP1, and 40% of gRAD51C/D positive patients. There was no difference in the rate of biallelic loss by age in patients with gPVs in HRD genes. Biallelic inactivation in tumors in patients with low-penetrance CRC-risk gPVs was 40% versus 14% in patients with low-penetrance non—CRC gPVs (p = < 0.0001). Conclusions: Germline-somatic integration implicated nearly half of all identified gPVs in CRC carcinogenesis. While the highest rate of biallelic inactivation was seen in high/moderate penetrance CRC-risk genes, biallelic inactivation in non-CRC genes, such as the HRD genes, warrants further assessment to discern the contribution of these gPVs to CRC carcinogenesis. Our findings may have implications for novel treatment strategies exploiting these pathways.