Out of 253 patients fulfilling criteria for a first episode of schizophrenic illness, 120 entered a randomised placebo-controlled trial of maintenance neuroleptic medication on discharge; they were followed to relapse or loss to follow-up, for two years or to the end of the study. Of those on active medication, 46% relapsed, as did 62% of those on placebo; the most important determinant of relapse was duration of illness prior to starting neuroleptic medication. This finding might be because extended duration of symptoms before admission is more likely to be present in illnesses which in any case will have poor prognosis, or because susceptibility to relapse is reduced by early institution of treatment. The study provides no data on which a decision between these alternative explanations can be based.