A framework is identified for modeling credit risk in agriculture. A CreditRisk+ type model is deemed most suitable for agricultural lending. The CreditRisk+ model is modified to overcome its drawbacks by incorporating recent research that accounts for sector correlations and uses a more stable and accurate algorithm. The model is applied to AgStar Financial Services, ACA, a cooperative agricultural lender, in order to determine how such a lender may adapt this model for portfolio risk analysis and to make capital and portfolio management decisions. The model generates a loan loss distribution, which is used to derive the lender’s expected and unexpected losses for the overall portfolio and individual loans. The model shows that AgStar is more than adequately capitalized based on the parameters estimated using 1997-2002 data. Since AgStar’s capital position is lower than that of most other associations, this raises the issue of overcapitalization within the Farm Credit System.