Urine testing does not show whether a person is impaired from the use of marijuana. Urine testing only shows THC metabolites, which are chemical grandchildren of marijuana use. When a person smokes marijuana, the chemicals released in the smoke travel through the blood to various receptors in the nervous system. The interaction of these chemicals with these receptors produce the various effects of marijuana intoxication: muscle relaxation, appetite stimulation, reduction of pain, certain cognitive effects., etc.

As the chemicals start to break down in the bloodstream, they become metabolities. Metabolites do not interact with the nervous system receptors. They are often stored in fatty tissue of the body and then they are slowly released through the urine and feces as waste. Urine testing measures how many of these waste metabolites are in the urine.

Urine testing shows that a person has used marijuana sometime in the last three weeks. For heavy users, metabolites may linger slightly longer than three weeks before they are entirely eliminated from the body. For light users, metabolites may not be present in the urine beyond a week from the time of use. Many factors affect the length of time a person’s body will hold metabolites: a person’s weight, level of activity, fluid intake, marijuana use frequency and intensity, and the person’s natural metabolism.

Urine metabolites of THC do not appear in the urine immediately upon smoking marijuana. It takes a few hours for the body to process THC to the point that it is entering the body’s waste system. Therefore a urine test of someone who just smoked but who did not smoke any before would produce a negative result. This person could be quite impaired. Therefore a urine test for impairment is not meaningful.

Blood tests of whole blood or blood serum/plasma do measure the immediate level of THC in the body. A person who has just smoked marijuana will experience intoxication almost immediately. The level of THC in the blood reflects the level of intoxication. However, in the case of testing for marijuana impairment in the situation of a driving accident for example, officers at the scene are not equipped to administer a blood test. By the time a person is taken to the police station or other location where a blood test can be administered, the THC levels in the blood may have dropped significantly.

In an effort to reduce the number of drivers who may put the public at risk through impaired driving, some states have developed “zero tolerance” laws which say that any level of THC metabolites in the body (urine or blood) constitute evidence of impaired driving. Such laws unfairly assume that a person who smoked marijuana on a previous day or even two weeks prior should be considered intoxicated.

Here are a few key points:

– Most studies find no effect of marijuana on psychomotor functions after four hours. This suggests that a waiting period of about three hours after smoking a medium to strong social dose (15-20 mg) will be sufficient to reduce a driver’s impairment to that comparable to a blood alcohol content of less than 0.03%. 

– A blood alcohol content of 0.08% corresponds approximately to a THC concentration in serum of 9-10 ng/mL. (nanograms per milliliter – a nanogram is one-billionth of a gram)

– Persons who consumed marijuana within the last hour are likely impaired and will typically produce a blood serum THC concentration of greater than 20 ng/mL

– Persons who consumed marijuana 48 hours earlier or who were exposed to second hand smoke will likely show a THC/blood serum level of 0-2 ng/mL.<http://www.norml.org/index.cfm?Group_ID=6492>

– Variable measurements of THC in identical blood samples among various forensic laboratories suggest that a safety margin of standard deviation should be included in testing standards. Researchers suggest a standard deviation of 3-4 ng/mL for blood serum testing when finding a concentration of 10 ng/mL or higher.

– Effects-based laws are the most reliable means of determining whether a driver is impaired. Effects-based laws require states to prove that a driver was “under the influence” of marijuana in light of the totality of facts in the case: driver’s conduct, testimony of witnesses, field sobriety tests, as well as any blood or urine tests. Such laws allow for cases where a regular marijuana user may produce a high blood level of THC but is not necessarily impaired.

From The Drug Testing Legal Manual by Kevin Zeese:
Drug Testing Legal Manual and Practice AIDS, by Kevin B. Zeese.  2nd Edition June 1996: West Group.

Arthur McBay, a foremost forensic toxicologist who served as the Chief Medical Examiner for the state of North Carolina states:

"A finding of 20ug L of THC in plasman (10ug/L in blood) probably indicates that marijuana was smoked with the hour and with 10 ug/L iun plasma within two hours. THC concentrations greater than 50 ug L could indicate smoking within 20 minutes. Concentrations of THC-COOH (the THC metabolite) of 10 ug L in plasma are attained in less than 10 minutes after beginning smoking and remain of 10 ug/L for 6 hours. They will exceed those of THC in 20 to 30 minutes.. . The intensity of 'high' increases rapidly for the first 10 minutes of smoking, peaks in the next 20 minutes and is reported to last for up to 4 hours after smoking started. It is unlikely that a range of plasma THC concentrations could be reliably equated with impaired performance. This differs from the way THC concentrations peak and rapidly decline but more closely resembles the curve of THC-COOH concentrations versus time. It is possible that elevated concentrations of THC and THC-COOH remain in chronic heavy smokers for many hours. A group of smokers was reported to have serum concentrations of 2 to 8 ug/L of THC and 27 to 93 ug/L of THC-COOH despite that fact that they were instructed to refrain from marijuana usage for at least 24 hours before testing."

In 1994 the Forensic Drug Abuse Advisor reported on research conducted by the Addiction Research Center which demonstrates that estimating the time of marijuana ingestion and determining impariment are becoming more accurate. They report that peak levels of marijuana occur in the blood even before subjects finish smoking. See, Estimating Time of Marijuana Ingestion and Degree of Impairment, 6 Forensic Drug Abuse Advisor 29 (1994) citing Cone et al "Relating blood concentrations to THC and metabolites to pharmacologic effects and time of marijuana usage," 15 Therapeutic Drug Monitoring 527 (1993). In order to see impaired effects the general rule is that THC levels must be higher than 10 ng/ml. Estimating the time of marijuana use is accurately done by mathematical  formulas comparing THC levels with THC metabolite levels. Two models have been reported that provide 95% confidence in this determination. See Huestis et al, 16 Journal of Analytic Toxicology 283-290 (1992).