Drought has very complex spatiotemporal evolution process and is highly impacted by various factors. Drought indices or indicators assimilated from only one drought-related variable are typically insufficient in characterizing different drought types at multiple scales. To overcome this problem, composite drought indexes via integrating multi-source information have been proposed in recent years. This study reviews the current status of composite drought indexes and summarizes three commonly utilized methods for constructing the indexes including weight combination method, multivariable joint distribution function method, and machine-learning-based modelling method. Moreover, we introduce the applications of composite drought indicators in the operational drought monitoring services in the United States, Europe, and China. Further research should focus on the mechanism of drought propagation, multi-source data merging, and the evaluation of composite drought indexes.