Is radon-222 ingestion a settled environmental exposure pathway?

Radon-222 ingestion from drinking water is commonly treated as a minor exposure pathway relative to inhalation of radon released into indoor air. This interpretation derives largely from the modelling framework consolidated by the U.S. National Academy of Sciences (NAS), which estimated that ingestion contributes approximately 10% of the total radon-in-water dose. However, the NAS explicitly recognised that the ingestion risk partition is governed by a poorly constrained biological parameter — the fraction of radon absorbed across the gastric wall prior to exhalation — and noted that plausible values could span up to two orders of magnitude. Because this parameter directly determines the ingestion dose coefficient, the widely cited 10% contribution represents a central estimate within a broad admissible range rather than a quantitatively constrained value. Despite this, the 90/10 partition has been embedded in subsequent international guidance without empirical narrowing of the governing biokinetic parameter. In parallel, radon-222 is excluded from total indicative dose screening frameworks for drinking water, reinforcing its marginalisation within ingestion-based assessments. The ingestion coefficient is therefore operationally stabilised in regulatory practice, while its quantitative basis remains sensitive to unresolved biological assumptions. This technical note argues that radon ingestion should not be regarded as a quantitatively resolved exposure pathway from an environmental radioactivity perspective. We show that (i) earlier assessments assigning a larger ingestion contribution remain compatible with the uncertainty bounds acknowledged by the NAS; (ii) standard ingestion models treat in-vivo decay and downstream progeny production as negligible under the same biokinetic assumptions that govern radon absorption and clearance; and (iii) radon dissolved in water functions within the uranium-238 decay chain as a transfer mechanism linking short-lived and longer-lived contributors to ingestion dose. Together, these considerations indicate that the apparent stability of the inhalation–ingestion partition reflects modelling continuity and regulatory convention more than empirical constraint. • The 90/10 inhalation–ingestion split for radon-222 originates from a single model. • Foundational assessments acknowledged two orders of magnitude uncertainty in ingestion dose. • Standard ingestion models truncate the decay chain by assumption, not physical necessity. • Radon-222 ingestion remains uncertainty-dominated and model-dependent in practice. • The marginalisation of radon ingestion reflects modelling lineage, not empirical closure.