In the last few years, calls for marijuana to be researched as a medical therapy have increased. It may be time for us to consider the same for psychedelic drugs. Two general classes of such drugs exist, and they include LSD, psilocybin, mescaline and ecstasy (MDMA).

All are illegal in the United States because they carry a high risk of abuse. They can also cause harm. The best-known adverse event is persistent flashbacks, though these are believed to be rare. More common are symptoms like increased heart rate and blood pressure, anxiety and panic.

Some people have pointed to anecdotal evidence of positive effects. Ayelet Waldman, a novelist and former federal public defender, wrote a memoir about how microdosing of the drugs turned her life around. But these drugs — like all drugs — carry risks that should not be ignored. With plenty of potential downsides, and no proven upsides, it’s not surprising that such drugs have been shunned.

In recent years, however, research has begun to show promise in treating a number of ailments.

People with life-threatening illnesses can also suffer from anxiety, which is hard to treat, especially when patients are on many other medications. In 2014, a small randomized controlled trial was published that examined if LSD could be used to improve this anxiety. The treatment included two LSD-assisted psychotherapy sessions conducted two to three weeks apart. Anxiety was significantly reduced in the intervention group for up to a year. Such results, however, could have been due in part to a placebo effect.

More common are studies of the use of psychedelics to treat abuse or addiction to other substances. A 2012 meta-analysis of studies exploring LSD’s potential to treat alcoholism looked at six randomized controlled trials. They included more than 500 patients, with follow-up of three to 12 months. The interventions usually involved one dose of LSD, given in a supervised setting, coupled with therapy. Alcohol use and misuse were significantly reduced in the LSD group for six months; differences seemed to disappear by one year. Similar studies using psilocybin have also shown promising results.

There was an open label study — meaning there’s no placebo or attempt to mask treatment information — of three doses of psilocybin as part of a tobacco cessation program. It found that 12 of 15 participants (who had smoked an average of more than 30 years) remained abstinent six months after the program began and 16 weeks after their last treatment. That’s a much higher rate than seen in traditional programs to help people quit smoking.

Other uses might exist as well. Researchers examined the potential for MDMA in the treatment of chronic and treatment-resistant post-traumatic stress disorder. At two months after therapy, more than 80 percent of those in the treatment group saw a clinical improvement versus only 25 percent of those in the placebo group. These researchers later followed up with participants in the study, and found that the beneficial effects lasted for at least four years, even with no further treatment with psychedelics. Similar studies have also seen improvements in symptom scores.

As with marijuana, though, studies like these are the exception, not the rule. It is very, very difficult to do research on psychedelic compounds because they, like pot, are classified as Schedule I controlled substances, meaning they have a very high potential for misuse and no accepted uses. Schedule II drugs also have a high potential for abuse, but are considered to have potential benefits. These include OxyContin, fentanyl, Percocet and even opium.

To engage in research in Schedule I drugs, scientists have to get approval from the Drug Enforcement Administration. To obtain a license, research labs must have inspections to prove that they are capable of storing the drugs and protecting them from misuse. In Britain, the added costs of licensing and security can cost a lab about £5,000 a year, or nearly $6,500. Unfortunately, the costs in the United States are not as well documented.

Because of this, much of the research on these drugs is old; a lot of it took place before the United States and other countries categorized these drugs in the 1960s. What research has occurred since has often taken place in countries that are more permissive in their experiments.

Given the potential dangers inherent in these drugs, it’s important to stress that research would need to be closely monitored. Although the drugs are relatively safe compared with substances like heroin or cocaine, and aren’t nearly as addicting, they still pose psychological and physical risks.

People with a family or personal history of psychotic or psychiatric disorders should be particularly wary, and perhaps be excluded from trials. Research requires safety monitoring, careful planning and significant support throughout. We need to watch carefully for adverse outcomes, both expected and unexpected. We need to make sure protocols are transparent and reproducible.

We also need to acknowledge that we need more research before anyone attempts to use these drugs as medicine. They’re typically coupled with professional therapy in studies, and we still aren’t sure there are benefits.

But it may be time to time to reconsider our current classification of controlled substances. Clearly we must continue to be vigilant about whether drugs pose physical harm to patients. But we could assess drugs using additional measurements, including the potential for dependence; social costs through damaged family and social life; and financial costs through health care, social care and the need for police involvement.

Using these metrics, it’s hard to argue that alcohol and tobacco should be legal for adults while marijuana and psychedelics should be considered so dangerous they’re hard to study. Likewise, opioids are considered widely acceptable in practice, yet appear to do far more harm.

With the potential to help curb more serious addictions and ease the symptoms of mental illnesses, it seems odd to continue to make it nearly impossible to research the therapeutic potential of psychedelics.

(This article originally appeared on nytimes.com and was written by Aaron Carroll)