We present MIDAS (Memory-Efficient I/O-Improved Drainage Analysis System), a fast, cross-platform toolkit for analyzing water flow and drainage patterns. Designed for high-performance parallel computing, MIDAS includes three core algorithms for flow accumulation, watershed delineation, and longest flow path analysis. It offers multiple interfaces---GRASS addons, a QGIS plugin, Python and R bindings, and a command-line tool---making it easy to use in both desktop and scientific environments. MIDAS efficiently handles large terrain datasets with low memory use and delivers significant speedups compared to traditional tools.