Droughts continue to cause significant damage across most societies despite maximizing reservoir efficiency. Reservoirs often fail to control drought impacts due to lack of reliable drought information that considers both water supply and demand. This study introduces a novel reservoir drought index (RDI) based on the reservoir sizing method to consider water supply and demand in drought analysis. The drought risks were analyzed using paleo and instrumental data alongside different demand scenarios to understand the role of demand changes on drought risks. RDI effectively represents actual reservoir storage compared to the dam inflow indicator. Paleo data adds value by revealing more severe drought events (71% more severe RDI, 20% higher severity, and 20% longer duration) and the most critical type of events (high stress, low endurance, and long duration) not shown in instrumental results. Water demand is crucial in controlling drought risks. Even minor changes in water demand (5 percentile level changes in demand) can increase drought risks by over 20%. The risks increase more for extreme events rather than moderate ones. Notably, increasing demand complicates extreme event analysis as multiple event get merged into a few mega-drought events. The suggested RDI and results addressing the role of water demand in drought can provide robust information for designing drought planning and mitigation strategies.