Drought is a natural phenomenon that has environmental and socio-economical drawbacks. Especially in arid and semi-arid regions, human activities are closely linked to the water supply and agricultural water use. Although the consequences of drought are prolonged, immediate actions are needed in practice which urges the continuous need for drought monitoring. The present study addresses a regional frequency analysis (RFA) for extreme drought events including severity, duration, and magnitude over Iran. Standardized precipitation index (SPI) time series with 1, 3, 6, 9, and 12 moving averages are determined from 106 meteorological stations for the period 1993-2016. Using Ward’s clustering analysis, the drought characteristics are grouped into different clusters and their homogeneity is confirmed by the heterogeneity measure test based on the L-moment approach. The results of RFA indicate that both generalized Pareto (GP) and Pearson type 3 (PE3) distribution functions are the best-fitted regional models to the most identified homogenous clusters of all three drought characteristics, by which the quantiles of each drought characteristic related to different return periods, T = 2, 5, 10, 25, 50, 75, and 100 years, are estimated. The spatial pattern of the drought characteristics for all SPI time scales shows that extreme droughts in terms of severity, duration, and magnitude may occur everywhere in the country regardless of local climate conditions. As such, even humid and rainy regions including northern, northwestern, and western parts of Iran that receive high annual precipitation would encounter extreme and severe drought characteristics. It is concluded that the drought risks in the region are mostly the outcome of mismanagement, water resource allocation, and agricultural water use but could be exacerbated due to climatic events.