Does THC Dilate Pupils? What Controlled Pupillography Found
Police reports often describe pupil size and reaction to light as signs of drug influence. Those observations may sound objective, but the scientific question is more exact: what was measured, under what lighting, at what time, with what instrument, and compared with what baseline?
A 2025 prospective, placebo-controlled study in BMC Ophthalmology used automated pupillography to measure the effects of intravenous delta-9-THC in healthy volunteers who did not regularly use cannabis. Contrary to the generalized belief that THC necessarily dilates the pupils, the study found smaller pupils after THC administration. It also found a dampened pupillary light reflex.
The result is important, but it must be kept inside the study’s limits. The experiment used intravenous THC, a small sample, and controlled equipment. It did not validate unaided roadside pupil estimates as a stand-alone test of cannabis impairment.
What pupillography measures
The pupil changes size in response to light through coordinated neurological and muscular activity. Pupillography records those changes under controlled conditions and can quantify features that an unaided observer cannot reliably estimate.
The study evaluated several measures:
- Baseline pupil diameter before the light stimulus
- Relative amplitude of constriction
- Absolute contraction amplitude
- Time required for constriction
- Latency before the response began
- Velocity of contraction
These are different biological features. A pupil can be smaller at baseline while also reacting more slowly or less fully to light. Compressing all of them into a phrase such as “pupils were abnormal” discards the details needed for interpretation.
What the controlled study did
Fifteen volunteers received an intravenous bolus of THC and four received placebo. Participants were cannabis-naive or had abstained for at least one month. The volunteers were monitored in a controlled hospital setting, and pupil measurements were collected before dosing and for five hours afterward.
The study was prospective and placebo controlled, but it was not randomized, double-blind, or a crossover design. The volunteers were blinded; study personnel were not. The first pupillography measurement after dosing occurred at 20 minutes because early blood collection occupied the initial period and some volunteers were unable to cooperate with eye testing during the strongest early effects.
Those design facts matter. The experiment offered far more measurement control than an ordinary roadside encounter, but the small and unequal groups limit how broadly the results can be generalized.
THC reduced pupil size in this experiment
The researchers reported that pupil diameter decreased after THC. They described the result as miosis rather than mydriasis.
That finding conflicts with the common categorical assertion that THC causes dilated pupils. The paper also noted that earlier research had been inconsistent: some studies reported increased pupil diameter, others reported decreased diameter, and another found no significant change.
The scientifically defensible conclusion is not that THC always constricts pupils. It is that pupil dilation is not a universal or necessary consequence of THC exposure, and this controlled intravenous study observed constriction.
The pupillary light reflex was dampened
The relative amplitude of pupil constriction decreased significantly after THC compared with placebo. Constriction time and contraction amplitude were also reduced. Latency and contraction velocity changed in the reported direction but did not reach statistical significance for those individual comparisons.
In plain language, the pupil’s response to light was less pronounced, and some aspects of the response were slower or shorter. The authors concluded that pupillography objectively detected a physical effect of THC on the eye.
That is not the same as proving unsafe driving. A measurable physiological effect establishes that something changed under the experimental conditions. Translating that change into a legal or functional conclusion requires evidence connecting the measurement to the person’s actual ability to operate a vehicle.
Route of administration limits roadside generalization
The volunteers received THC intravenously. Most real-world cannabis use involves inhalation or oral ingestion. Intravenous delivery avoids variable absorption and allows controlled pharmacokinetic modeling, but it is not the ordinary route encountered in driving investigations.
The route changes how quickly THC enters the blood, how concentrations rise and fall, and when subjective or physical effects occur. The researchers themselves identified intravenous administration as a limitation.
A roadside opinion should therefore not treat this study as a direct lookup table for a smoked product, a vaporizer, or an edible. It is evidence that THC can alter pupil size and light-reflex dynamics, not a universal schedule for every route and user.
Visual estimation is not automated pupillography
Pupillography uses standardized illumination, timing, camera-based measurement, and numerical output. A roadside pupil estimate may be affected by ambient light, flashlight distance, adaptation time, officer position, measurement card placement, contact lenses, eye color, stress, fatigue, medication, and medical conditions.
The controlled study does not establish that an officer can reproduce the same measurements by sight. Nor does it establish that a single pupil-size category identifies THC as the cause. Opioids, neurological conditions, ophthalmic drugs, autonomic state, and many ordinary variables can alter pupil size or reaction.
The larger limitations of field-test interpretation are discussed in What the Field Sobriety Tests Actually Measure.
Presence, physical effect, and impairment are different propositions
A blood result can support the presence of THC in the specimen. Pupillography may demonstrate a contemporaneous physical effect on the eye. Neither proposition automatically proves that the person was unable to drive safely.
The study authors suggested combining concentration evidence with a physical-effect measurement. That is an interpretation proposed by the researchers. The study did not establish a validated pupillography threshold that determines driving impairment for every person.
This distinction parallels the issue explained in Concentration Is Not Impairment. The evidence must be connected to timing, function, and the individual rather than treated as a self-proving label.
What a case-specific review should obtain
A meaningful review of pupil evidence should obtain:
- Video showing the full eye examination, instructions, lighting, timing, and repeated observations.
- Ambient-light measurements or documentation of the lighting environment.
- The method used to estimate or measure pupil diameter and the instrument’s calibration records.
- Whether both eyes were measured and whether the results were symmetrical.
- Adaptation time before each light stimulus and the flashlight distance and duration.
- Contact-lens use, recent ophthalmic examination, eye injury, surgery, disease, and visual complaints.
- Medications and substances capable of changing pupil size or light response.
- Neurological, autonomic, metabolic, pain, anxiety, sleep, and fatigue evidence.
- The exact driving, observation, examination, and blood-collection times.
- The claimed cannabis route, dose, product, use time, and prior-use pattern.
- The laboratory THC and metabolite data, method, validation, uncertainty, and reporting limits.
- The scientific basis for converting the observed pupil feature into an opinion about driving ability.
When alcohol or another drug is present, combined effects and alternative causes require separate analysis. See Cannabis and Alcohol Together: Impairment and Field Signs.
Important limits of the study
The THC group included 15 volunteers and the placebo group only four. The study was single-blind, nonrandomized, and not a within-subject crossover. Study personnel knew the treatment assignment.
Measurements began 20 minutes after administration, so the experiment may not describe the earliest pupil response near the initial concentration peak. Some measurements could not be completed because participants did not tolerate the bright-light stimulus, creating the possibility that the strongest responders were underrepresented at those times.
The study involved people who did not regularly use cannabis. Tolerance and habitual use may change the relationship. The authors also described increased light sensitivity and possible driving consequences as possibilities requiring further support, not as demonstrated outcomes of a driving experiment.
The bottom line
The categorical claim that THC necessarily causes dilated pupils is not supported by this controlled study. After intravenous THC, the volunteers showed smaller pupils and a dampened response to light on automated pupillography.
That does not make every small pupil a sign of cannabis, and it does not convert a measured light-reflex change into proof of unsafe driving. Route, timing, measurement method, baseline, alternative causes, and the complete behavioral evidence still control the interpretation.
Primary source
Kleine-Brueggeney M, Priemer F, Konietschke F, Theiler L, Greif R. Effects of intravenous delta-9-THC on pupillary reaction and pupil size: a prospective, placebo-controlled trial in healthy volunteers not regularly consuming cannabis. BMC Ophthalmology. 2025;25:286. PubMed record. PubMed Central full-text record. Springer Nature journal record.
Editorial note: This article paraphrases the published research. The source PDF, publisher layout, tables, and figures are not reproduced.
