Relative to a normal-state control group, subjects responding under the influence of hypnosis were significantly less accurate on both tasks. Heightened susceptibility to misleading implications was shown to be the major source of hypnotic inferiority. Neither hypnosis nor control groups revealed any relationship between confidence and accuracy. These results replicate and extend the findings of previous research. It is recommended that the use of hypnosis in criminal investigations be limited to eliciting testimony from witnesses with memory blocks and to be used only as a last resort. Tabular data and 19 references are provided. (Author abstract modified)