In a series of investigations, suppression of the auditory-evoked P50 potential to the second of two paired clicks presented 500 msec apart has been shown to be absent in schizophrenic patients, whereas normals suppress their second response to less than 20% of the first response. The phenomenon has been discussed as a possible trait marker for schizophrenia. The present study with 19 schizophrenics and 23 healthy control subjects was intended as an extended replication of the phenomenon using different stimulus parameters and a slightly different method of measuring P50 amplitudes. Replication was unsuccessful, revealing only weak suppression scores in normal subjects not significantly superior to schizophrenics. Retest sessions yielded generally stronger P50 suppression suggesting that the stability of the measure over time is questionable. The methodological changes are discussed as possible sources of this failure to replicate. It is concluded that the conditions under which P50-suppression occurs should be better clarified in order to facilitate replication. © 1990.