This dataset contains time-series measurements collected from a real-world water distribution network located in Stockholm, Sweden, and was used in the study “Leakage detection in water distribution networks using machine-learning strategies” (https://doi.org/10.2166/ws.2023.054) published in Water Science & Technology (IWA Publishing, 2023). The dataset includes operational and hydraulic variables acquired from multiple sensing points in the network, such as flow and pressure measurements, recorded at regular time intervals. The data represent both normal operating conditions and leakage scenarios, enabling the development, validation, and benchmarking of data-driven methods for leakage detection in water distribution systems. The dataset is particularly suited for research on: Leakage and anomaly detection in water distribution networks Time-series analysis and pattern recognition Machine learning and data-driven modeling for smart water systems Evaluation of centralized and distributed detection approaches All data are provided in their processed form as used in the referenced publication, ensuring reproducibility of the reported results. The dataset may also serve as a benchmark for future studies addressing reliability, robustness, and explainability of leakage detection methods under real operational conditions. Users of this dataset are kindly requested to cite the associated journal article when using the data in academic publications.