In November 1996 the people of California approved Proposition 215, an initiative that could, in effect, make marihuana legally available as a medicine in the United States for the first time in many years. Under the new law, patients or their primary caregivers who possess or cultivate marihuana for medical treatment recommended by a physician are exempted from criminal prosecution. The treatment may be for "cancer, anorexia, AIDS, chronic pain, spasticity, glaucoma, arthritis, migraine, or any other illness for which marihuana provides relief." Physicians may not be penalized in any way for making the recommendation,  which may be either written or oral.1

Although this law will provide a defense for patients in criminal cases, it is not a long-term solution for the problem it addresses, because the federal government will not allow authority over drugs to revert to the states. Nevertheless, it is likely that similar laws will be more widely adopted, because the California vote signals a growing public impatience with the  present legal obstacles to medical cannabis.* According to a 1995 poll conducted by the American Civil Liberties Union, 85% of Americans believe that marihuana should be available as a medicine.2

The history of cannabis as a medicine goes back at least 5,000 years to ancient China and extends well into the twentieth century. Nineteenth-century European and American physicians were familiar with marihuana. Between 1840 and 1900, more than one hundred papers on the therapeutic use of the drug then known as Indian hemp were published in medical journals. It was recommended as an appetite stimulant, muscle relaxant, analgesic, hypnotic, anticonvulsant, and treatment for opium addiction. As late as 1913, Sir William Osler cited it as the best remedy for migraine.3 In the nineteenth century medical cannabis was administered chiefly in oral form, as an alcoholic solution. But the potency of these solutions varied, and patients responded erratically to oral ingestion. Shortly after the turn of the century, synthetic alternatives became available for insomnia and moderate pain, two of the most common indications for the use of cannabis. In the United States, what remained of its medical use was effectively eliminated by the Marihuana Tax Act of 1937, which was ostensibly designed to prevent non-medical use but made cannabis so difficult to obtain that it was removed from the pharmacopeia. Since 1970 the federal government has classified it as a Schedule I drug, with a high potential for abuse, no accepted medical use, and lack of safety for use under medical supervision.

In 1972 the National Organization for the Reform of Marijuana Laws (NORML) entered a petition to transfer marihuana to Schedule II so that it could be legally prescribed. As the proceedings continued, other parties joined NORML, including the Physicians Association for AIDS Care. In 1986 the Drug Enforcement Administration (DEA) finally acceded to the demand for public hearings required by law. During the hearings many patients and physicians testified and thousands of pages of documentation were introduced. In 1988 the DEA's Administrative Law Judge, Francis L. Young, declared that marihuana in its natural form had a currently accepted use in treatment in the United States and therefore fulfilled the legal requirement for transfer to Schedule II. He noted in his opinion that it was "one of the safest therapeutically active substances known to man."4 His order was overruled by the DEA.

Since 1976 a few patients have been able to obtain medical marihuana legally through an Individual Treatment Investigational New Drug program (commonly referred to as a Compassionate IND) that was discontinued in 1992 after a deluge of applications from people with AIDS. About three dozen patients eventually received marihuana under the program, and eight survivors are still receiving it. Beginning in the 1970s, 35 states also enacted legislation establishing special IND research programs under which patients with certain disorders could use cannabis. These projects eventually proved unworkable and are now in abeyance. For all but eight residents of the United States, marihuana remains a forbidden medicine.

Shortly before the federal government blocked the last legal access, organizations were founded to distribute medical marihuana in open defiance of the law: the cannabis buyers' clubs. These clubs purchase cannabis wholesale and provide it to patients on the written recommendation of a physician. Patients are generally asked to donate enough to cover costs; those who cannot pay may receive marihuana free. Although some of the buyers' clubs have been raided and closed, most are still flourishing, and new ones are being organized. Backers of the California initiative have talked of legitimizing buyers' clubs through state action.

Marihuana is a strikingly safe, versatile, and potentially inexpensive medicine. When we reviewed its medical uses in 1993 after examining many patients and case histories, we were able to list the following: nausea and vomiting in cancer chemotherapy, the weight loss syndrome of AIDS, glaucoma, epilepsy, muscle spasms and chronic pain in multiple sclerosis, quadriplegia and other spastic disorders, migraine, severe pruritus, depression, and other mood disorders.5 Since then we have identified more than a dozen others, including asthma, insomnia, dystonia, scleroderma, Crohn's disease, diabetic gastroparesis, and terminal illness. The list is not exhaustive.6

Cannabis could be therapeutically significant even if only a few patient with a given symptom or disorder benefited. For example, several people told us that cannabis helped them free themselves of alcohol, opiate, or tobacco addiction. We do not know how common or how powerful this capacity is, but further exploration would be worthwhile even if only a few highly motivated substance abusers could benefit. As nineteenth-century physicians understood, anyone who is sophisticated about drugs would gladly exchange alcohol, opiates, or tobacco for marihuana.

Cannabis has also been found useful in the treatment of osteoarthritis. Aspirin is believed to cause more than 1,000 deaths annually in the United States. More than 7,600 annual deaths and 70,000 hospitalizations caused by non-steroidal antiinflammatory drugs (NSAIDs) are reported. Gastrointestinal complications of NSAIDs are the most commonly reported serious adverse drug reaction.7 Long-term acetaminophen use is thought to be one of the most common causes of end-stage renal disease.8 Cannabis smoked several times a day is often as effective as NSAIDs or acetaminophen in osteoarthritis, and there have been no reports of death from cannabis.

It is often objected that the evidence of marihuana's medical usefulness, although powerful, is merely anecdotal. It is true that there are no studies meeting the standards of the Food and Drug Administration, chiefly because legal, bureaucratic, and financial obstacles are constantly put in the way. The situation is ironical, since so much research has been done on marihuana, often in unsuccessful efforts to show health hazards and addictive potential, that we know more about it than about most prescription drugs. In any case, controlled studies can be misleading if the wrong patients are studied or the wrong doses are used, and idiosyncratic therapeutic responses can be obscured in group experiments.

Anecdotal evidence is the source of much of our knowledge of drugs. As Louis Lasagna has pointed out, controlled experiments were not needed to recognize the therapeutic potential of chloral hydrate, barbiturates, aspirin, insulin, or penicillin.9 Anecdotal evidence also revealed the usefulness of propranolol and chlorothiazide for hypertension, diazepam for status epilepticus, and imipramine for enuresis. All these drugs had originally been approved for other purposes.

In the experimental method known as the single patient randomized trial, active and placebo treatments are administered randomly in alternation or succession. The method is often used when large-scale controlled studies are inappropriate because the disorder is rare, the patient is atypical, or the response to treatment is idiosyncratic.10 Several patients have told us that they assured themselves of marihuana's effectiveness by carrying out such experiments on themselves, alternating periods of cannabis use with periods of abstention. We are certain that the medical reputation of cannabis is derived partly from similar experiments conducted by many other patients.

Some physicians may regard it as irresponsible to support, let alone advocate the medical use of cannabis on the basis of anecdotal evidence, which seems to count successes and ignore failures. That would be a serious problem only if cannabis were a dangerous drug. The years of effort devoted to showing that marihuana is exceedingly dangerous have proved the opposite. It is safer, with fewer serious side effects, than most prescription medicines, and far less addictive or subject to abuse than many drugs now used as muscle relaxants, hypnotics, and analgesics.11

Thus it can be argued that even if only a few patients could get relief from cannabis, it should be made available because the risks would be so small. For example, many patients with multiple sclerosis find that cannabis reduces their muscle spasms and pain. A physician may not be sure that such a patient will get better relief from marihuana than from the baclofen, dantrolene, and high doses of diazepam that the patient has been taking, but it is certain that a serious toxic reaction to marihuana is extremely unlikely, and risk-benefit considerations therefore make it worth trying.

The chief legitimate concern is the effect of smoking on the lungs. Many physicians find it difficult to endorse a smoked medicine. Although cannabis smoke carries even more tars and other particulate matter than tobacco smoke, the amount needed by most patients is extremely limited. Furthermore, when marihuana is an openly recognized medicine, solutions for this problem may be found, perhaps by the development of a technique for inhaling of cannabinoid vapors. Even today, the greatest danger of using marihuana medically is not impurities in the smoke but illegality, which imposes much unnecessary anxiety and expense on suffering people.

Furthermore, if cannabis were not prohibited, it would be less expensive than most conventional medications. The cost of medical marihuana would be $20 to $30 an ounce, or about 30 cents per cigarette. One cigarette usually relieves the nausea and vomiting produced by cancer chemotherapy. So does the standard dose of ondansetron (Zofran), the best present legally available treatment, which costs the patient $30 to $40 -- at least 100 times the price of marihuana.

A synthetic version of delta-9-tetrahydrocannabinol, the main active substance in cannabis, has been available in oral form for limited purposes as a Schedule II drug since 1985. This medicine, dronabinol (Marinol), is sometimes said to obviate the need for medical marihuana, but patients and physicians who are familiar with both disagree. A patient who is severely nauseated and constantly vomiting, for example, may find it almost impossible to keep a pill or capsule down. Oral THC is erratically and slowly absorbed into the bloodstream; the dose and duration of action of smoked marihuana are easier to titrate. Furthermore, oral THC occasionally makes many patients anxious and uncomfortable, possibly because cannabidiol, one of the many substances in marihuana, has an anxiolytic effect.12

When medical use of marihuana in the United States was effectively outlawed by the Marihuana Tax Act, the American Medical Association, to its credit, opposed the ban. Since then, unfortunately, physicians have become largely ignorant about marihuana. As both victims and agents, they have too often promoted the spread of misinformation and frightening myths. But now they are relearning what their nineteenth-century counterparts knew, and they are coming to this knowledge in an unusual way -- not from articles in medical journals or from pharmaceutical company advertisements, but from their patients. In a typical case a person with, say, HIV infection discovers that marihuana slows or even reverses his weight loss. On his next visit to the doctor he steps on the scale and proves it. Eventually the doctor's incredulity is overcome, and he may tell other patients.

This remarkable, perhaps unique learning process has now been going on for some time. In the spring of 1990 investigators randomly selected more than 2,000 members of the American Society of Clinical Oncology (one third of the membership) and mailed them an anonymous questionnaire. Almost half of the recipients responded and only 43% of them said the available legal antiemetic drugs (including oral synthetic THC) provided adequate relief to all or most of their patients. Forty-four percent had recommended the (illegal) use of marihuana to at least one patient, and half would prescribe it to some patients if it were legal. On average they considered smoked marihuana more effective than oral synthetic THC and roughly as safe.13

Many patients who use marihuana to relieve symptoms ranging from muscle spasms to severe depression have discussed with us the reactions of their physicians. A few physicians condemn the marihuana use, and some pretend to ignore it or profess indifference, but most offer some encouragement or moral support -- what would be termed a "recommendation" under the California initiative. Yet federal law declares marihuana to be unsafe for use even under medical supervision. Obviously physicians confronting the needs of their patients can recognize the foolishness of this law. But most have been either afraid to do more or unable to provide further help because they know too little.

Thus many physicians may now be asked to assume responsibilities for which they are unprepared. More and more patients will approach them with questions about marihuana, and they will have to provide answers. They must not only listen more carefully to their patients but educate themselves and one another. They will have to learn which symptoms and disorders may be treated better with marihuana than with conventional medications, and they may need to explain how to use marihuana. It is not as simple as taking a pill; some preparation and instruction may be required, both to attain therapeutic goals and to avoid unwanted effects. The psychoactive effects must be explained to marihuana-naive patients, who may otherwise suffer some degree of anxiety. Many patients will also have to be taught the mechanics of smoking and the correct way to titrate the dose.

Eventually physicians must acknowledge more openly that the present classification of marihuana as a Schedule I drug is scientifically, legally, and morally wrong. They have both a right and a duty to be skeptical about therapeutic claims, but only after dismissing fears and doubts connected with the stigma of illicit non-medical drug use. Advocates of medical marihuana are often charged with the hidden agenda of encouraging "recreational" use. This false accusation represents in distorted form a certain truth: some opponents of medical marihuana are interested only in insuring that the dangers of non-medical use are exaggerated and prohibition continues.

Pharmaceutical companies will not pursue the clinical research needed to test marihuana's therapeutic potential, because they have little to gain and much to lose. The federal government, so far, has only blocked the way. In 1994 Donald Abrams of the University of California, San Francisco, sought approval for a pilot study comparing smoked marihuana with oral THC in the treatment of the AIDS wasting syndrome. Although his project was approved by the FDA and several institutional review boards and advisory committees, the National Institute on Drug Abuse and the Drug Enforcement Administration prevented him from receiving the marihuana he needed.14

The protocol was finally accepted, but only after it had been changed from an efficacy to a safety study. The California vote should persuade federal authorities to relent. But even if they do, a research program designed to study the many clinical uses of this versatile drug will take years, and meanwhile other ways must be found to accommodate the needs of a rapidly growing number of patients.

*Arizona voters have approved a related but more restrictive initiative which permits physicians, "notwithstanding any law to the contrary", to prescribe Schedule I controlled substances, including marihuana, "to treat a disease, or to relieve the pain and suffering of seriously and terminally ill patients." The physician must document scientific research supporting the prescription and must obtain written approval from a second physician (Arizona Revised Statutes, 13-3412.01). Like the California law, this initiative may provide a criminal defense for patients. However, few physicians are likely to write such a prescription, since there is no legal way to fill it and they could still be charged with violating federal law.


1. California Health and Safety Code Section 11362.5

2. Belden N, Russonello H. American voters' opinions on the use and legalization of marijuana. Survey conducted for the American Civil Liberties Union, 1995 (unpublished).

3. Osler W. The principles and practice of medicine, 8th ed. New York: Appleton, 1913, p. 1089.

4. Drug Enforcement Administration (hereafter DEA), in the Matter of Marijuana Rescheduling Petition, Docket 86-22, Opinion, Recommended Ruling, Findings of Fact, Conclusions of Law, and Decision of Administrative Law Judge, September 6, 1988.

5. Grinspoon L, Bakalar JB. Marihuana, the Forbidden Medicine. New Haven: Yale University Press, 1993.

6. Grinspoon L, Bakalar JB. Marihuana, the Forbidden Medicine (revised and enlarged edition). New Haven: Yale University Press, in press, 1997.

7. Gurkirpal S, Ramey DR, Morfeld D, Shi H, Hatoum HT, and Fries, JF. Gastrointestinal tract complications of nonsteroidal anti-inflammatory drug treatment in rheumatoid arthritis. Archives of Internal Medicine 1996;156: 1530-6.

8. Perneger TV, Whelton P, and Klag MJ. Risk of kidney failure associated with the use of acetaminophen, aspirin, and nonsteroidal antiinflammatory drugs. N Engl J Med 1994;331:1675-9. Ronco PM, Falhault A. Drug-induced end-stage renal disease. Editorial, N Engl J Med 1994;331:1711-2.

9. Lasagna L. Clinical trials in the natural environment. In Drugs between Research and Regulations, ed. Stiechele C, Abshagen W, and Koch-Weser J. 1985. Darmstadt: Steinkopff Verlag, pp. 45-9.

10. Larson EB. N-of-1 clinical trials: a technique for improving medical therapeutics. Western Journal of Medicine 1990;152:52-6. Guyatt GH, Keller JL, Jaeschke R, et al. The N-of-1 randomized controlled trial: clinical usefulness. Annals of Internal Medicine 1990;112:293-9.

11. Grinspoon L, Bakalar JB. Marihuana, the Forbidden Medicine. New Haven: Yale University Press, 1993, pp. 137-54.

12. Chang AE, et al. Delta-9-tetrahydrocannabinol as an antiemetic in cancer patients receiving high-dose methotrexate: a prospective, randomized evaluation. Annals of Internal Medicine 1979;91:819-824. Zuardi AW, Shirakawa I, Finkelbarb E, and Karnio IG. Action of cannabidiol on the anxiety and other effects produced by delta-9-THC in normal subjects. Psychopharmacology 1976;76:245-50.

13. Doblin R, Kleiman M. Marihuana as anti-emetic medicine: a survey of oncologists' attitudes and experiences. Journal of Clinical Oncology 1991;9:1275-80.

14. Grinspoon L, Bakalar JB. Letter to the editor, N Engl J Med 1995;33:670-1. Abrams DI, Child CC, Mitchell TF. Letter to the editor, N Engl J Med 1995;33:671.