Lucid dreaming — the experience of becoming aware, mid-dream, that one is dreaming — has been reported for thousands of years, described in Tibetan Buddhist dream yoga, Aristotelian writings, and medieval mystical texts. For most of that history it sat in a difficult evidential position: intensely subjective, easy to fake, hard to verify from the outside. That changed in the 1970s.
In 1975, the British psychologist Keith Hearne, working at the University of Hull, was studying a self-described lucid dreamer named Alan Worsley. The two agreed in advance that if Worsley became lucid inside a REM dream, he would signal by making a specific pattern of left-right eye movements — deliberately, from within the dream. Worsley's regular dream eye movements would be indistinguishable from normal REM activity, but the pre-agreed pattern would stand out. On the night of 12 April 1975, Hearne's polysomnograph recorded exactly that pattern during a confirmed REM period. It was, arguably, the first objective evidence of consciousness inside a dream.
Independently, and unaware of Hearne's work, the American psychophysiologist Stephen LaBerge at Stanford developed a similar paradigm and published his results in the early 1980s. LaBerge and colleagues extended the technique to show that lucid dreamers could signal not only awareness but also the timing of internal events: they could estimate ten seconds in-dream and signal at the start and end, and the intervals matched real elapsed time closely. They could hold their breath in the dream, and the corresponding respiratory pauses appeared in the polysomnograph. Singing and counting in the dream produced different patterns of hemispheric activation on EEG, mirroring waking neural signatures.
More recent work has pushed the boundary further. In 2021, a multi-lab collaboration led by Karen Konkoly and Ken Paller published a study in *Current Biology* showing two-way communication with lucid dreamers during REM sleep. Researchers spoke simple math questions ("what is eight minus six?") to sleeping participants, who answered from inside the dream using eye movements or facial muscle twitches. Not every trial succeeded, but enough did — across four independent labs, in four languages — to establish that real-time interactive communication with a dreaming person is possible.
Practically, lucid dreaming can be trained. The best-studied induction techniques include *reality checking* (building a daytime habit of asking, "am I dreaming?" and doing a check that fails in dreams — reading a line of text twice, trying to push a finger through the palm), *MILD* (mnemonic induction, in which you rehearse the intention to notice you are dreaming as you fall asleep), and *WBTB* (wake-back-to-bed, in which you sleep for around five hours, wake briefly, then return to sleep during the REM-rich final period of the night). None of these is magic. Studies typically find that a few weeks of consistent practice produces modest but real increases in lucid-dream frequency, with wide individual variability.
Lucidity has been proposed as a therapeutic tool for chronic nightmares, particularly in PTSD populations — the theory being that a dreamer who realizes she is dreaming can change the course of the nightmare. Early trials are promising but not yet definitive, and lucid-dream therapy is not currently a first-line treatment. Anyone with recurrent trauma nightmares deserves proper clinical care, not a self-taught technique.
What is well established is the phenomenon itself. Lucidity is real, measurable, and trainable. What lucid dreams *mean*, what they can accomplish, and how far the state can be developed are much more open questions — and, honestly, some of the most interesting ones left in the field.