Human health risks associated with groundwater fluoride have been assessed using USEPA method in few parts of India, but those assessments were conducted based on a single value for each parameter, which may lead to certain inaccuracy and uncertainties in results. In this study, a higher degree of accuracy in health risk assessment was achieved through Monte Carlo simulations, sensitivity analysis and uncertainty analysis. As fluoride hazards to human health are chronic, a total of 4560 water samples (N = 4560) were collected during consecutive four seasons (2 pre-monsoon and 2 post-monsoon seasons; 1140 samples/season) from the entire Birbhum district, covering all the blocks and geological settings to obtain the spatiotemporal variation of fluoride level. The Empirical Bayesian Kriging geostatistical model was employed to determine fluoride endemic areas. Amongst all blocks, Nalhati-1 had exhibited the highest fluoride level (18.25 mg/L). The study revealed that most of the blocks are vulnerable to groundwater fluoride due to its occurrences in excess level. The average and 95th percentile values of total hazard index indicate that the infants and children populations of the district are more susceptible than the adults and teens. The sensitivity analysis revealed that water ingestion rate and fluoride concentration are the most influential parameters for higher risk of fluoride-related health hazards. Health risks were evaluated through ingestion and dermal exposure routes for infants, children, teens, and adult residents. The exposure was much higher through ingestion than dermal contacts. Ingestion of defluoridated water will reduce the health risks associated with groundwater fluoride exposure.