The CDC estimates that 29.1 million Americans have type 2 diabetes, which is currently the seventh leading cause of death in the US. It has been proposed that disordered glucagon regulation is a major contributor to the diabetic phenotype. We hypothesized that variants in five genes in the glucagon pathway are associated with response of diabetes-related glucose, insulin, and body composition traits to three antidiabetic interventions in the DPP and that rare coding variants would be enriched compared to other populations and associated with type 2 diabetes and diabetes-related traits in the Old Order Amish (OOA). We have discovered an association between the GCG variant, rs5649, that is predicted to modify the consensus splice site, with increased baseline-adjusted fasting glucose at one year in the metformin and placebo groups but not in the other two treatment groups, suggesting that it may change the structure of glucagon and decrease the ability of metformin to reduce glucagon signaling. We have performed the only known study of a PCSK2 coding variant common in any population. We found this variant, R430W, and the GCGR variant D458H to be twice as common among OOA individuals with diabetes as those with normal or impaired glucose tolerance, although this difference is not statistically significant. In addition, we examined the relationship between the GCGR coding variant G40S and type 2 diabetes risk that has been previously reported and did not see an association despite being better powered than previous studies because of the increased allele frequency of 14% in the OOA population. Finally, we described the only reported GCGR G40S homozygotes, none of whom have type 2 diabetes. In sum, we have provided data on the effects of variants in five genes in the glucagon pathway that can be used in future work to understand the contribution of the glucagon pathway to type 2 diabetes and to inform the development of antidiabetic treatments targeting this pathway.