Can Breath Alcohol Peak More Than an Hour After the Last Drink?

Answer: Yes. Breath alcohol concentration (BrAC) can continue rising after a person finishes drinking, and there is no single post-drink interval that proves everyone has reached a declining concentration. In a 2026 supervised social-drinking study of 42 adults, the interval from the final drink to the highest sampled BrAC ranged from 23 minutes before the last drink to 76 minutes afterward; the median was 13 minutes afterward. Most peaks occurred relatively soon after drinking ended, but several profiles were delayed or irregular. These observations describe measured breath alcohol time courses—not blood alcohol concentrations, an individual’s condition at an earlier event, or proof of impairment.

The primary evidence is O’Connor, Wang, and Vargas, “Variability in breath alcohol concentration profiles during controlled social drinking” (Journal of Analytical Toxicology, 2026). It is an observational analysis of repeated breath measurements obtained during supervised drinking; it does not establish a universal waiting period for forensic examinations.

Can breath alcohol peak more than an hour after the last drink?

It did in this study. The longest reported interval between the final drink and the highest measured BrAC was 76 minutes. The mean interval was 17.1 minutes, with a median of 13 minutes and a range of −23 to +76 minutes. A negative interval means the participant’s highest sampled concentration occurred while drinking was still underway. The authors observed that approximately three quarters of participants peaked within roughly 30 minutes after finishing, yet a minority peaked more than an hour later.

The distinction between an observed peak and an individual’s precise physiological maximum matters. Breath samples were collected approximately every 30 minutes, not continuously. The actual highest concentration could have occurred between collections. The largest observed delay in this 42-person dataset is neither a population-wide upper bound nor a rule that applies to a particular person.

What did researchers actually measure?

Forty-two adults attended supervised social-drinking sessions at California State University, Los Angeles. They consumed mixed drinks prepared from 80-proof spirits, selected mixers, and had food available without a standardized meal. They drank at differing paces within a staff-guided dosing plan. Drinking sessions lasted 45 to 172 minutes, and the mean consumption was 4.9 standard drinks (range 2.5–8.0) under the study’s definition of approximately 14 grams of ethanol per drink.

Researchers collected baseline and approximately 30-minute interval breath samples using an Intox EC/IR II instrument. Their sequence included a gas-standard suitability check, blanks, a breath sample, a further blank, and a replicate breath sample; only successful sequences entered the analysis. The study reported BrAC in grams of ethanol per 210 liters of breath (g/210 L). It did not draw serial venous blood to establish contemporaneous blood alcohol concentrations.

Measured observations were used to calculate several additional descriptors: the largest sampled BrAC (Cmax), the time of that peak relative to the final drink (Tmax), and the slope of a fitted regression line through selected post-peak data. The regression-derived slope is a model estimate from breath observations, not another directly measured biological analyte.

What did the 42-person study find?

Source: primary study methods, Table 2, Figures 3–4, discussion, and DOI 10.1093/jat/bkag038. Numerical means and ranges describe the study participants and should not be applied as universal thresholds.

Why can measured breath alcohol keep rising after drinking stops?

The end of beverage consumption and the end of systemic alcohol absorption are different events. Ethanol can still be moving from the gastrointestinal tract into circulation after the final sip. At the same time, the body is distributing and eliminating alcohol. The net concentration-time profile reflects these overlapping processes, not a switch that flips at the moment the drink is finished.

The researchers intentionally permitted variation in food intake, mixers, and drinking pace to capture patterns closer to social drinking than a single fixed-dose experiment. Individual curves included an early peak, a delayed peak, and an example with more than one local maximum. Such curves underscore uncertainty about the point at which an individual transitions from rising to reliably falling BrAC. They do not establish which physiological factor caused each person’s pattern: the study did not experimentally isolate food, gastric emptying, mixers, and all other candidate determinants one at a time.

Even a generally descending sequence can show small irregularities. A flat or irregular interval might reflect sampling variation, ongoing absorption, physiology, or other causes. An expert should not label a particular interval “elimination only” merely because a later observation is lower, without considering the timing, sequence, and quality of the relevant measurements.

Is a measured breath concentration the same as a measured blood alcohol concentration?

No. BrAC is quantified from exhaled breath using the instrument’s measurement system and reporting convention; blood alcohol testing measures ethanol in a sampled blood specimen through a separate analytical procedure. In this experiment, the time series consists of breath results, expressed as g/210 L. Venous blood was not sampled to test how an individual person’s blood results tracked the breath readings.

Accordingly, a rising BrAC curve in this study supports a claim about observed breath measurements over time. It does not directly demonstrate a rising venous BAC in a particular defendant, patient, or employee. A statutory or assumed blood-to-breath conversion cannot create blood measurements that the researchers did not collect.

For background on how breath measurements differ from blood laboratory tests, see the site’s verified Breath Alcohol Measurement and Blood Alcohol Testing by Headspace GC-FID guides. Those are method primers, not replacements for case-specific records.

Does the mean elimination slope justify retrograde BAC calculation?

Not by itself. The investigators estimated each participant’s declining breath concentration slope by fitting a line to a selected post-peak segment. The mean slope was 0.0186 g/210 L per hour, but the participant values ranged from 0.0134 to 0.0261 in those same units. A mean describes this group. It cannot identify which rate applied to a different person at a different time.

A retrograde estimate also depends on where the measured event falls relative to absorption completion. Extrapolating backward with a post-peak decline rate through a period when alcohol was still being absorbed can misrepresent the earlier time course. The study provides evidence against assuming a uniform post-drink transition time; it does not invalidate every scientifically constrained calculation. Rather, it supports choosing a model only when its prerequisites are actually supported.

Any event-time interpretation should keep separate the observed specimen, exact sample time, timing and quantity of drinking, analytical and sampling uncertainty, and the assumptions required to project backward. The study did not reconstruct the BrAC or BAC of any outside case.

What records matter in a dispute about a rising-alcohol theory?

The scientifically relevant first question is whether the proposed time sequence is sufficiently documented to test the theory. Important records include the time of each drink, how consumption was estimated, the elapsed interval to each breath or blood collection, the specimen and unit reported, the instrument or laboratory method, and the presence of repeat measurements. A single later result alone cannot reveal an individual peak time.

For breath instruments, the sequence and quality information matter: observation and collection conditions, blank and standard checks where applicable, replicate results, error messages, and the reported units. For blood analyses, specimen collection and laboratory records answer a different measurement question. The two matrices should not be merged without stating the model and its uncertainty.

See also the site’s verified Pharmacokinetics in Forensic Toxicology guide for the general relationship among absorption, distribution, metabolism, and elimination. The 2026 study adds a narrow empirical point about when the largest sampled breath value occurred relative to the last drink.

What does this study not prove?

It is a controlled but intentionally flexible social-drinking investigation, not an on-road driving trial, an impairment-validation study, or a controlled comparison of matched blood and breath samples. Participants received individually guided doses toward a breath target; drinking amounts and durations differed. Food intake was not standardized, demographic data were incomplete for some variables, and a sample of 42 adults cannot establish universal limits. Approximate 30-minute sampling also limits time resolution for identifying a true physiological peak.

The study measured no driving performance, mental-state capacity, crash causation, or observed impairment of a person in a forensic case. It cannot prove either that an accused individual was impaired earlier or that an individual was not impaired earlier. A later positive breath result is a measurement at its collection time; an earlier concentration or a legal or medical conclusion demands additional evidence.

Frequently asked questions

How long after the last drink can BrAC rise?

In the 42-person 2026 experiment, the largest recorded breath result occurred as late as 76 minutes after the last drink. The sample’s median was 13 minutes. This is an observed study range, not a guaranteed maximum or a personal prediction.

Does BrAC always peak after the final drink?

No. The reported interval extended down to −23 minutes: some participants’ largest collected breath result occurred before the final drink. Whether a measured series is rising depends on the actual timing and values.

If a breath test occurs an hour after drinking, is absorption definitely finished?

No. The study included delayed peaks beyond an hour, so a fixed one-hour rule would not describe everyone in this sample. The study does not establish a different universally valid replacement interval.

Is the highest measured breath sample the exact biological peak?

Not necessarily. Sampling occurred approximately every 30 minutes. The maximum among those measurements may miss a higher or earlier concentration between collections.

Can a breath study establish an individual venous BAC curve?

No. The study did not obtain paired blood concentrations. BrAC and blood BAC are different specimen measurements, and an individual blood curve cannot be directly recovered from breath observations alone.

Did the researchers calculate alcohol elimination from measured breath results?

Yes. They fitted post-peak breath measurements with regression lines to estimate decline rates. The slopes were derived quantities, not direct measurements of liver metabolic activity or individual blood alcohol elimination.

Does a rising BrAC prove that someone was sober or unimpaired at an earlier time?

No. A rising concentration pattern may affect an estimate of an earlier concentration, but this research did not measure event-specific driving behavior or impairment. Earlier concentration and earlier performance remain different questions.

Scientific takeaway

O’Connor and colleagues provide direct evidence that the interval between the last drink and the largest observed breath alcohol result varies substantially under supervised social-drinking conditions. The central interpretation is limited but important: do not assume that absorption is finished at a fixed number of minutes after the final drink, and do not treat a breath-based population mean as an individualized blood concentration or impairment conclusion. Sound forensic analysis starts with the actual measurement, specimen, sampling times, and defensible assumptions.

Primary scientific reference

O’Connor RS, Wang J, Vargas JR. Variability in breath alcohol concentration profiles during controlled social drinking. Journal of Analytical Toxicology. 2026;50(6):bkag038. https://doi.org/10.1093/jat/bkag038.