A placebo is a fake pill with no medicine in it. People thought it worked only when they believed it was real medicine.

But in 2010, doctors at Harvard told patients with stomach problems, "This is a fake pill." The patients took it twice a day for three weeks. Many of them still felt better than patients who took nothing. Scientists are still studying why.

The placebo effect happens when a pill without any active ingredient makes people feel better. It was long believed that this only works if patients think the pill is real. In 2010, however, Harvard researchers gave 80 patients with irritable bowel syndrome either no pills or pills that were clearly described as placebos.

The patients were told the pills contained no medicine but might still help the body heal itself. After three weeks, about 60% of the placebo group said their symptoms were adequately relieved, compared with about 35% of the no-pill group. Later studies of back pain found similar results. However, because patients knew what they were taking, the studies could not be blinded.

Placebos have traditionally been thought to work only through deception: patients improve because they believe an inert pill is real medicine. A 2010 Harvard trial challenged this assumption. Researchers divided 80 patients with irritable bowel syndrome into two groups. One received pills openly described as containing no active ingredient, along with an explanation that placebo effects might trigger the body's self-healing processes; the other received no pills but equally attentive care.

After three weeks, the open-label placebo group reported greater global improvement, and about 60% felt adequate relief, compared with about 35% of controls. A 2016 trial in 97 patients with chronic low back pain found reduced pain and disability, and a 2021 meta-analysis of 11 trials reported a moderate overall effect.

Several explanations have been proposed, including the ritual of taking a pill, conditioning from past experiences with medicine, and a supportive relationship with clinicians. Important limitations remain, however. Because patients know what they are taking, the trials cannot be blinded, outcomes are mostly self-reported, and results vary widely. Placebos mainly ease subjective symptoms such as pain or fatigue; they do not shrink tumors or kill bacteria, so they should never replace proper treatment.