Eighty Ss were first tested for base-level response to a pain-producing stimulus and then were re-tested on the same pain stimulus after receiving 1 of 8 experimental treatments. The 8 treatments were arranged in a 2 x 2 x 2 factorial design: presence or absence of hypnotic induction procedure; presence or absence of instructions for anesthesia; and presence or absence of demands for honest reports. Neither the hypnotic-induction procedure nor the demands for honesty affected the Ss'reports of the degree of pain experienced. The anesthesia instructions--"think of the hand as numb and insensitive as if it were a piece of rubber..."--produced an equal degree of pain reduction in hypnotic and non-hypnotic Ss and in Ss who were and those who were not exposed to demands for honesty. The results indicate that (a) Ss' reports of pain are less affected by demands for honesty and are more closely related to their actual experiences than has been previously assumed and (b) instructions which direct Ss to exercise cognitive control over painful sensory input are effective (with or without 'hypnosis') in reducing the experience of pain.