Assessment of agricultural drought risk at critical times before and during the growing season can provide sufficient time to policy makers and farmers to implement appropriate strategies to reduce the potential of risk. The purpose followed in the present study is to develop a statistical model to estimate the quality of agricultural drought risk for rainfed barley crop (prior to planting and as well during the growing season) in East Azarbaijan Province, Iran. Model input variables include the weekly Standardized Precipitation Index (SPI) figures within various time windows with the only output variable being the risk of crop yield (with two groups namely: high vs. low risk). A multivariate technique was employed to model the relationship between input vs. output variables during different growth stages. The results revealed that the calibrated model can be used to assess the real-time agricultural drought risk for barley crop at pre-planting and as well during crop growth stages by retaining the previous, and adding the current WSPI data as the crop passes through the various developmental stages. The accuracy of risk assessment increases as the crop barley develops. Model validation revealed that the most appropriate time to asses drought risk for barley crop within the study area is the flowering stage, with the results for the crop at tillering stage also applicable.