This study utilizes comparisons and Probit regression analysis to determine the influence of previous migrations and other variables on the likelihood of future migrations of agricultural loan credit risk. The Farm Credit System association data set contains a large number of lender risk-rated agricultural loans. The lender risk ratings used are less likely to migrate than ratings based on credit score proxies. The results indicate that the direction of previous migrations significantly influences future migrations in a trendreversal pattern. Forecasting future migrations remains difficult even though the marginal effect of a previous migration on the likelihood of a future migration is quite large.