What the calculator estimates
The calculator produces a broad earlier-to-later detection window and places the entered elapsed time into one of three educational bands: higher modeled likelihood, uncertain/transition range, or lower modeled likelihood. It does not calculate a laboratory-grade EtG concentration.
Inputs used
- Standard-drink count: grouped into exposure bands rather than treated as perfectly precise.
- Hours since the last drink: compared with the modeled range.
- Cutoff: 100, 200, or 500 ng/mL changes the breadth of the window.
- Session length: longer sessions modestly widen the range.
- Repeated drinking: adds uncertainty and widens the upper range.
Why we do not show an “exact EtG level”
Urine EtG is affected by dose, absorption, timing of urine voids, hydration and creatinine concentration, assay characteristics, renal function, and individual biology. A simple half-life equation can look scientific while implying accuracy the available inputs cannot support—see our calculation formula guide for exactly where that false precision creeps in, and our half-life calculator for a version that reports a range instead.
Our bands are calibrated to patterns reported in controlled drinking and outpatient studies, then deliberately widened to avoid presenting a group average as an individual prediction.
Primary research used
| Primary study | Design / population | Cutoffs or window | How it informs the site |
|---|---|---|---|
| Jatlow et al., 2014 | Controlled dose-ranging alcohol challenge | EtG and EtS over time | Dose response, detection limits, and interpretation uncertainty |
| Lowe et al., 2015 | Alcohol-dependent outpatients | 100–500 ng/mL; 24–120 h | Cutoff sensitivity and specificity across days |
| McDonell et al., 2015 | Outpatient self-report comparison | One to five days | Light versus heavy drinking detection by cutoff |
| van de Luitgaarden et al., 2019 | Population cohort | 100 and 500 ng/mL | Population-level sensitivity/specificity context |
Model logic and versioning
Version 1.0 groups standard-drink exposure into five broad bands, applies a cutoff multiplier, then widens the upper range for repeated drinking and longer sessions. The displayed band compares elapsed time with that modeled range. These deliberately conservative heuristics communicate relative factors; they are not fitted to predict a numeric laboratory concentration.
The model source is visible in the calculator interface and is updated only when a change can be explained against primary evidence. We do not train the model on visitor entries because those entries are not collected.
Limitations: The model has not been clinically validated as a diagnostic device. It cannot account for all health conditions, sample validity, laboratory methods, or incidental exposure. It must not be used to make medical, legal, employment, treatment, or supervision decisions.
Updates and corrections
Research sources were last reviewed on July 29, 2026. To suggest a correction or additional primary source, use our corrections process or email hello@etgcalculatorpro.com.