Health hazard risk mapping (HHRM) is an important technique used to estimate the potential health risk of an individual, a group, or an entire community of a region. To further progress this work, 67 samples were collected through field investigation in the dry season i.e., from March to early June of 2021 from different parts of hardy rock dominated Purulia district. In this study, 14-health hazard causative factors were considered such as Depth (m), pH, EC (μS/cm), HCO 3 (mg/L), As (μg/l), Ca 2+ (mg/L), Cl − (mg/L), F − (mg/L), K + (mg/L), Mg 2+ (mg/L), Na + (mg/L), NO 3 ( mg/L), PO 4 2− ( mg/L), SO 4 2− ( mg/L). All of these parameters are selected using multi-collinearity and Pearson’s correlation test. Furthermore, three important machine learning algorithms namely bagging, random forest (RF), and an ensemble of bagging and RF were employed to assess the HHRM. The outcome of the learning models were evaluated by statistical validating methods such as AUC-ROC, sensitivity, specificity, accuracy, precision, F-score, kappa, and Taylor diagram. The result of validating techniques ensure that ensemble technique is more reliable in training (AUCROC-0.934, sensitivity-0.917, specificity-0.925, accuracy-0.921, precision-0.925, F-score-0.922 and kappa-0.851) and validating dataset (AUCROC-0.911, sensitivity-0.904, specificity-0.905, accuracy-0.902, precision-0.907, F-score-0.907 and kappa-0.819) with Taylor diagram ( r = 0.94) followed by bagging and RF. The produced result shows the central part of the study area especially the districts of Bagmundi, Balarampur, Arsha, Purulia I and II, Raghunathpur are significantly susceptible to the health hazard due to poor water quality that covers around 15% of the total area.