Scientists believe they have located the part of the brain where people's dreams are created.
The researchers studied the patient's brainwaves as she slept
A team from the University Hospital of Zurich, Switzerland, made the discovery after treating a woman who stopped dreaming after she had a stroke.
It had affected an area deep in the back of the brain - and they suggest this is the area controlling dreaming.
The researchers, writing in the Annals of Neurology, say the finding offers a new focus for dream research.
The 73-year-old patient lost a number of brain functions, mostly related to vision, with her stroke.
Most came back after a few days - but she then stopped dreaming. Before her stroke, she had dreamt three or four times a week.
The loss of the ability to dream - along with visual disturbances - following damage to a specific part of the brain, is called Charcot-Wilbrand syndrome, named after the eminent neurologists Jean-Martin Charcot and Hermann Wilbrand, who first described it in the 1880s.
The syndrome is quite rare, especially cases that lack symptoms other than dream loss.
The Swiss researchers decided to monitor the patient to try to discover which part of the brain was affected in people with the condition.
They monitored the woman's brainwaves for six weeks as she slept.
Her sleep was not disrupted, and she continued to have REM (rapid eye movement) sleep as normal.
This is significant, because dreaming and REM sleep occur together, although research has pointed to different brain systems underlying the two.
The researchers say their findings appear to confirm that dreaming and REM sleep are driven by independent brain systems.
Scans of the patient's brain showed the stroke had damaged areas located deep in the back half of her brain.
Other studies have shown that some of this region is involved in the visual processing of faces and landmarks, as well as the processing of emotions and visual memories, a logical set of functions for a brain area that would generate or control dreams.
After around a year, the patient did begin to have occasional dreams, but no more than one per week.
She reported that her dreams were less vivid and intense than they were before the stroke.
Writing in the Annals of Neurology, Dr Claudio Bassetti, of the Department of Neurology at the University Hospital of Zurich in Switzerland, who led the research, said: "How dreams are generated, and what purpose they might serve, are completely open questions at this point.
"These results describe for the first time in detail the extent of lesion necessary to produce loss of dreaming in the absence of other neurological deficits.
"As such, they offer a target for further study of the localisation of dreaming."
He added: "Further conclusions about this brain area and its role in dreams will require more studies analysing dream changes in patients with brain damage."