Spatial aggregation can influence drought assessments where gridded data intersect irregular coastlines. This study quantified the sensitivity of drought indicators to three aggregation rules across five Ionian Islands, Greece. Monthly CHIRPS v3 precipitation and TerraClimate potential evapotranspiration for 1981–2024 were aggregated using cell–island intersection-area weighting, equal weighting of all valid intersecting cells, and equal weighting of valid cells with native-grid centers inside island boundaries. Source-grid geometry and full-period data availability were verified without interpolation or gap filling. The Standardized Precipitation Evapotranspiration Index (SPEI) was calculated at 3- and 12-month scales. Log-logistic distributions were fitted separately for each island, rule, scale, and calendar month using unbiased probability-weighted moments and a common 1981–2020 calibration period. Comparisons used area weighting as the reference and evaluated continuous differences and drought classification at SPEI ≤ −1. Across island–scale combinations, mean absolute SPEI differences were 0.022–0.055 for equal weighting and 0.005–0.018 for center-based selection. Among 5,215 valid island–scale–month records per comparison, drought classifications differed in 45 (0.86%) and 18 (0.35%) records, respectively. Among records classified as dry by either compared method, disagreement was 4.30% and 1.75%, respectively. The largest island–scale disagreement was 9/526 months (1.71%) for equal weighting in Ithaca at SPEI-3. For these products, islands, timescales, and separately calibrated series, drought classification was largely stable to aggregation choice but retained threshold sensitivity. The findings quantify methodological robustness, not independently validated accuracy or the general superiority of one rule. Small differences in standardized indices do not imply identical absolute hydroclimatic estimates.