Obtaining reliable estimates of insurance premiums is a critical step in risk sharing and risk transfer necessary to ensure solvency and continuity in crop insurance programs. Challenges encountered in the estimation include dealing with aggregation bias from using county level yield averages as well as properly accounting for spatial and temporal heterogeneity. In this study, we associate some of these challenges as classical small area estimation (SAE) problems. We employ a hierarchical Bayes (HB) SAE to obtain design consistent expected county level yields and Group Risk Plan (GRP) premiums for corm farms in Illinois using quasi-simulated data. Preliminary results show little bias (< 10%) in estimated expected county yields in several counties investigated. We found wide variation in GRP, APH and basis risk across counties for similar level of coverage and scale. Results show that farmers could lower their GRP premiums by as much as 30% by carefully choosing a coverage level and scale combination.