Updated arsenic toxicity values substantially alter health-risk interpretation in a regulation-compliant, geothermally influenced basin-fed drinking-water system.

Compliance with drinking-water guideline values is commonly treated as evidence of negligible health risk, yet guideline values and toxicological reference doses are revised on independent schedules. This study tested whether these two bases for judging water safety diverge in a basin-fed, geothermally influenced supply system, using paired municipal, spring, and river water samples (n = 42) from ten districts of Eskişehir Province, Türkiye. Seventeen trace elements were quantified by ICP-MS, and health risk was assessed deterministically and probabilistically using the January 2025 USEPA IRIS toxicological update for arsenic. Arsenic exceeded the WHO and Turkish national (TS266) guideline value of 10 µg/L in nearly a third of samples, yet the Water Quality Index classified almost all samples as "Excellent" under the same guideline values. Applying the updated arsenic toxicity values raised the Hazard Index and cancer risk above acceptable thresholds at most sites, whereas the legacy (1995) values flagged far fewer of the same sites as unacceptable on identical concentrations - a divergence attributable solely to the toxicological revision and confirmed robust to exposure-parameter uncertainty by Monte Carlo simulation. Arsenic, correlated with antimony, lithium, and boron, dominated cumulative risk, consistent with a geogenic, geothermally influenced source rather than treatment-related contamination; municipal and spring samples showed broadly similar trace-element profiles. These findings reveal a structural disconnect between regulatory compliance and contemporary toxicological evidence - not a flaw in the guideline values themselves - with implications for drinking-water risk assessment beyond arsenic and beyond this study area.