DENVER >> Angeline Chilton says she can’t drive unless she smokes pot. The suburban Denver woman says she’d never get behind the wheel right after smoking, but she does use medical marijuana twice a day to ease tremors caused by multiple sclerosis that previously left her homebound.

"I don’t drink and drive, and I don’t smoke and drive," she said. "But my body is completely saturated with THC."

Her case underscores a problem that no one’s sure how to solve: How do you tell if someone is too stoned to drive?

States that allow medical marijuana have grappled with determining impairment levels for years. And voters in Colorado and Washington state will decide this fall whether to legalize the drug for recreational use, bringing a new urgency to the issue.

A Denver marijuana advocate says officials are scrambling for limits in part because more drivers acknowledge using the drug.

"The explosion of medical marijuana patients has led to a lot of drivers sticking the (marijuana) card in law enforcement’s face, saying, ‘You can’t do anything to me, I’m legal,’" said Sean McAllister, a lawyer who defends people charged with driving under the influence of marijuana.

It’s not that simple. Driving while impaired by any drug is illegal in all states.

But it highlights the challenges law enforcement officers face using old tools to try to fix a new problem. Most convictions for drugged driving now are based on police observations, followed later by a blood test.

Authorities envision a legal threshold for pot that would be comparable to the blood-alcohol standard used to determine drunken driving.

But unlike alcohol, marijuana stays in the blood long after the high wears off a few hours after use, and there is no quick test to determine someone’s level of impairment — not that scientists haven’t been working on it.

Dr. Marilyn Huestis of the National Institute on Drug Abuse, a government research lab, says that soon there will be a saliva test to detect recent marijuana use.

But government officials say that doesn’t address the question of impairment.

"I’ll be dead — and so will lots of other people — from old age, before we know the impairment levels" for marijuana and other drugs, said White House drug czar Gil Kerlikowske.

Authorities recognize the need for a solution. Marijuana causes dizziness, slowed reaction time and drivers are more likely to drift and swerve while they’re high.

Dr. Bob DuPont, president of the Institute for Behavior and Health, a non-government institute that works to reduce drug abuse, says research proves "the terrible carnage out there on the roads caused by marijuana."

One recent review of several studies of pot smoking and car accidents suggested that driving after smoking marijuana might almost double the risk of being in a serious or fatal crash.

And a recent nationwide census of fatal traffic accidents showed that while deadly crashes have declined in recent years, the percentage of mortally wounded drivers who later tested positive for drugs rose 18 percent between 2005 and 2011.

DuPont, drug czar for Presidents Richard Nixon and Gerald Ford, wrote a paper last year on drugged driving for the Obama administration, which has made the issue a priority.

Physicians say that while many tests can show whether someone has recently used pot, it’s more difficult to pinpoint impairment at any certain time.

Urine and blood tests are better at showing whether someone used the drug in the past — which is why employers and probation officers use them. But determining current impairment is far trickier.

"There’s no sure answer to that question," said Dr. Guohua Li, a Columbia University researcher who reviewed marijuana use and motor vehicle crashes last year.

His survey linked pot use to crash risk, but pointed out wide research gaps. Scientists do not have conclusive data to link marijuana dosing to accident likelihood; whether it matters if the drug is smoked or eaten; or how pot interacts with other drugs.

The limited data has prompted a furious debate.

Proposed solutions include setting limits on the amount of the main psychoactive chemical in marijuana, THC, that drivers can have in their blood. But THC limits to determine impairment are not widely agreed upon.

Two states place the standard at 2 nanograms per milliliter of blood. Others have zero tolerance policies. And Colorado and Washington state are debating a threshold of 5 nanograms.

Such an attempt failed the Colorado Legislature last year, amid opposition from Republicans and Democrats. State officials then set up a task force to settle the question — and the panel couldn’t agree.

This year, Colorado lawmakers are debating a similar measure, but its sponsors concede they don’t know whether the "driving while high" bill will pass.

In Washington state, the ballot measure on marijuana legalization includes a 5 nanogram THC limit.

The measure’s backers say polling indicates such a driving limit could be crucial to winning public support for legalization.

"Voters were very concerned about impaired driving," said Alison Holcomb, campaign director for Washington’s legalization measure.

Holcomb also pointed to a failed marijuana legalization proposal in California two years ago that did not include a driving THC limit.

The White House, which has a goal of reducing drugged driving by 10 percent in the next three years, wants states to set a blood-level standard upon which to base convictions, but has not said what that limit should be.

Administration officials insist marijuana should remain illegal, and Kerlikowske called it a "bogus argument" to say any legal level of THC in a driver is safe.

But several factors can skew THC blood tests, including age, gender, weight and frequency of marijuana use. Also, THC can remain in the system weeks after a user sobers up, leading to the anxiety shared by many in the 16 medical marijuana states: They could be at risk for a positive test at any time, whether they had recently used the drug or not.

A Colorado state forensic toxicologist testified recently that "5 nanograms is more than fair" to determine intoxication. But, for now the blood test proposals remain politically fraught, with supporters and opponents of marijuana legalization hinging support on the issue.

Huestis, of the government-funded drug abuse institute, says an easy-to-use roadside saliva test that can determine recent marijuana use — as opposed to long-ago pot use — is in final testing stages and will be ready for police use soon.

Researchers envision a day when marijuana tests are as common in police cars as Breathalyzers.

Until then, lawmakers will consider measures such as Colorado’s marijuana DUI proposal, which marijuana activists say imperils drivers who frequently use the drug such as Chilton, the multiple sclerosis patient.

She says that since she began using pot she has started driving again and for the first time in five years has landed a job.

Chilton worries Colorado’s proposal jeopardizes her newfound freedom.