.    This following is an excerpt from "Psi Notes", prepared by William Braud,
Ph.D., of the Mind Science Foundation in San Antonio, Texas.

Question:  What percentage of a person's dreams are precognitive (foretell the
future) and how can we recognize the difference between a precognitive dream
and an ordinary dream?

Answer:  A large proportion of precognitive experiences occur during dreams. 
One survey indicates that as many as 65 percent of precognitive experiences
occurred during sleep.  Precognitive dreams also seem to provide more complete
and more accurate information than do waking psychic experiences.

.   There's no way to know with certainty what percentage of our dreams are
precognitive.  The content of the majority of our dreams is probably quite
mundane, involving replays of experiences of the day, perhaps some wish
fulfillment, and maybe even "random" content.  But now and then, dreamers do
have accurate glimpses of the future as they sleep.

.   The only way to know with certainty which dreams are precognitive and which
are not is to keep a dream diary of all dreams and check to see which come true
and which don't.  Some persons are able to associate certain feelings of
confidence in connection with psychic dreams - but these are very subtle
feelings which are difficult to put into words and which may differ from person
to person.

.   Let me describe a program of research in which we are more certain about
what's going on.  This research program was initiated by a New York
psychiatrist, Dr. Montague Ullman, as a result of his observation that he and
his patients were sharing telepathic dreams in the context of psychotherapy.  A
dream laboratory was set up at Maimonides Medical Center in Brooklyn.  Ullman,
along with his associates Stanley Krippner and Charles Honorton, designed
experiments in which persons spent the night in the dream lab.  They were
monitored electro-physiologically in order to detect physiological indications
of dreaming - these indications include:  an activated EEG, rapid eye
movements, and reduced muscle tension.  When these indications of dreaming
occurred, the sleeper was awakened and asked to describe his dream.  These
descriptions were tape-recorded and later transcribed.  The next day, a target
experience was randomly selected and the subject then went through some waking
sensory experience.  What was discovered was that the sleeper was able to have
accurate dreams about events of which no one was as yet aware at the time of
the dream, but which were randomly selected the next day.