Forensika › Articles

Forensic medicine · alcohol

The Widmark calculation in practice: back-calculation, balance and breath

Ján Šikuta, MD, PhD, forensic pathologist · Published 5 November 2026

“How much did he have in his blood at the time of the accident?” is one of the questions a forensic pathologist hears most often. The answer is never a single figure but a range, and its width depends on what is really known about the drinking and about the person.

One equation, three questions

The Widmark equation links three quantities: the amount of ethanol consumed, body mass multiplied by the factor r, and the elimination rate β. The concentration at time t is c = A / (m · r) − β · t. The same equation answers three different questions that are often mixed up in expert reports:

  • curve: what concentration a person could reach after a given drinking pattern,
  • back-calculation: what the concentration was at the time of the incident, given a later sample,
  • balance: how much alcohol must have been consumed to reach the measured concentration, and whether that fits the person's account.

The ethanol in a drink is volume × alcohol content × 0.789 (density of ethanol in g/mL). A half-litre of 4 % beer thus contains about 15.8 g of ethanol, four of them about 63 g.

The factor r: a fixed value or calculated from height and weight

Practice long relied on fixed conventional values of 0.7 for men and 0.6 for women. The factor r, however, expresses the ratio of total body water to blood water, and that varies considerably between people. Seidl et al. (2000) measured it in 529 adults: from 0.44 to 0.80 in women and from 0.60 to 0.87 in men. They derived regression equations from height and weight, and in drinking experiments the individual r matched measured values better than the fixed 0.6 and 0.7. When height is known, an individual r is a sensible choice; for a tall slim person and for an obese person a fixed factor shifts the result in opposite directions.

Back-calculation and its conditions

Back-calculation adds what was eliminated between the incident and the sample: c₀ = c + β · Δt. Example: a sample taken two hours after an accident showed 0.90 g/kg. With β of 0.12 to 0.20 g/kg/h the concentration at the time of the accident was 1.14 to 1.30 g/kg. The result is reported as a range, not as a single mid-value. For a conclusion in favour of the accused, the appropriate extreme value is chosen according to context.

The calculation is only meaningful if the person was in the elimination phase for the whole time between the incident and the sample. If they drank shortly before or after the incident, alcohol was still being absorbed and a linear addition alone overestimates the concentration, so a balance deduction has to be added. The 2025 statement of the Slovak Society of Forensic Medicine (SSLS) allows a simple back-calculation at a blood concentration of at least 0.21 g/kg (breath 0.10 mg/L), provided no more than 15 hours passed between the end of drinking and the examination. With a longer interval the uncertainty grows even when these conditions are met.

Uncertainty: a range, not a figure

Every input carries its own uncertainty: the amount and type of drink from a statement, body weight, r and β. Searle (2015) showed that the uncertainty cannot be expressed as one fixed percentage, as has sometimes been done in the literature, but has to be calculated for the case at hand. In practice this means reporting the range from the extreme values of β and, in a balance calculation, allowing for the absorption deficit, the part of the alcohol that never reaches the blood. According to the SSLS it is always at least 10 %, up to 20 % for low-alcohol drinks with a lot of food and rarely up to 30 %.

Breath and blood

The ratio of blood to breath alcohol is not a constant. Jones and Andersson (1996) measured it in 799 drivers at a mean of 2407 (95 % limits of agreement 1981 to 2833) after correcting for elimination between the samples. Statutory conversions in many countries use a lower ratio of 2000 or 2100, so for most people they work in the driver's favour: in this series only 4.3 % had a corrected ratio below 2100. In Slovakia, according to the SSLS statement of 22 January 2025, a breath result in mg/L is converted to g/kg directly with a factor of 2.1.

Differences between breath and blood also have practical causes: exhalation technique, the time between the breath test and blood sampling, and the quality of the analysis. In a paper I co-authored (Bauer et al., 2015), we drew attention above all to the human factor in analysis and recommended two mutually independent methods or the analysis of two biological materials.

What the SSLS statement (2025) says

These are professional thresholds for expert assessment in Slovakia. They are not statutory limits and give nobody permission to drive after drinking. What counts as an offence is set by law and decided by the competent authority or court.

  • up to and including 0.30 g/kg in blood and 0.14 mg/L in breath the concentration is insignificant for road traffic and other activities, laboratory error included,
  • at 0.80 g/kg in blood or 0.38 mg/L in breath and above nobody, not even an above-average person, is fit to take part in road traffic without endangering themselves or others,
  • blood is analysed by HS-GC-FID (or HS-GC-MS) and the result is the mean of at least two determinations,
  • the degree of impairment cannot be assessed from the urine concentration alone,
  • β is usually taken as 0.12 to 0.20 g/kg/h; below 0.30 g/kg or with serious liver damage it may fall to 0.10 to 0.12, and with prolonged heavy drinking above 2.00 g/kg it may exceptionally rise to 0.25.

After death different rules apply

In post-mortem samples ethanol can be produced by microorganisms, especially with decomposition, and central blood may gain alcohol by diffusion from the stomach. Kugelberg and Jones (2007) therefore recommend analysing vitreous humour as well, which is better protected from these effects, and the urine to blood ratio can indicate whether alcohol was still being absorbed at the time of death.

How 4n6 Tools handles it

The Ethanol: calculations tool in 4n6 Tools computes the curve, the back-calculation and the balance, offers an individual r after Seidl, the β range of the SSLS statement and breath conversion with the factor 2.1. It reports the result as a range, warns when the elimination-phase conditions are not met and, with two or more samples, estimates an individual β from the slope of the curve.

Related app: 4n6 Tools. This article is a professional overview for qualified experts; it does not replace a textbook, guidelines or an expert opinion.

References

  1. Widmark EMP. Principles and Applications of Medicolegal Alcohol Determination. Davis: Biomedical Publications; 1981.
  2. Seidl S, Jensen U, Alt A. The calculation of blood ethanol concentrations in males and females. Int J Legal Med. 2000;114(1–2):71–77. doi:10.1007/s004140000154
  3. Searle J. Alcohol calculations and their uncertainty. Med Sci Law. 2015;55(1):58–64. doi:10.1177/0025802414524385
  4. Jones AW, Andersson L. Variability of the blood/breath alcohol ratio in drinking drivers. J Forensic Sci. 1996;41(6):916–921.
  5. Kugelberg FC, Jones AW. Interpreting results of ethanol analysis in postmortem specimens: a review of the literature. Forensic Sci Int. 2007;165(1):10–29. doi:10.1016/j.forsciint.2006.05.004
  6. Bauer M, Bauerová J, Šikuta J, Šidlo J. Confrontation of knowledge on alcohol concentration in blood and in exhaled air. Soud Lek. 2015;60(2):21–24 (English abstract).
  7. Slovak Society of Forensic Medicine (SSLS). Position statement on assessing the influence of ethanol on a person. Bratislava, 22 January 2025.