The study focused on establishing the most efficient geographic insurance units (GIUs), as basis in the design of weather index-based crop insurance (WIBI) for rainfed rice in selected provinces in the Philippines, namely: Isabela, Iloilo and Bukidnon. The study aimed to characterize the said locations based on weather patterns, and crop risk profiles in terms of probability of crop failure in order to determine which areas will be subjected to similar index insurance designs, and which farms will be considered with a different one. Initial GIU designs were established based on political boundaries (municipal, district, or provincial) for operational considerations. Optimal GIU designs for each province considered were then identified using the two criteria: (1) area-specific weather distribution through mean monthly rainfall patterns; and (2) crop risk profiles based on the probability of crop failure due to rainfall deficit. Analyses were based on forty-one years of synthetically generated, and validated rainfall data from 2010 to 2050. In order to define the optimal GIUs for each province, comparison of the two sets of results were done based on their data requirements and practical efficiency of implementation in the field. For Isabela province, a total of twelve (12) GIUs were defined. In the case of Iloilo, seven (7) GIUs were defined. While for Bukidnon, five (5) GIUs were developed. Results of the study are intended to be used to minimize risk, and thus, reduce the premium for the insurance product. By defining GIUs for WIBI, the insurance products will have high potential in effectively helping farmers cope with and eventually manage climate risks.