Twenty patients meeting DSM-III criteria for Schizophrenic Disorder, Paranoid Type were studied. After a 30 day drug-free period the patients were randomized in two groups. During a year, patients in Group 1 received intramuscular injections of a placebo while Group 2 received recombinant IFN alpha-2b. Both groups took anti-psychotic medication (APM) as needed on an individual basis, depending on their psychiatric symptomatology. Double blind evaluations, were performed at the beginning and at the end of the trial, using the Brief Psychiatric Rating Scale (BPRS) and the Reyes Scale for Social Evolution (RSSE). Information about relapses was gathered such as months without relapse, number and duration of the relapses and maximum dosages of APM given for relapse control. The statistical analysis of the results was performed with a matched pairs sign or Student's t tests for comparisons of each group before and after treatment. Groups were compared between them using the Fisher's exact test for frequencies and Student's t test for continuous variables. In Group 1: only one patient improved on the BPRS score; two had improved ratings on the RSSE; 2 patients got worse; and there were no changes in the rest of the group; three patients had no relapses and one increased in relapse frequency. These changes were not significant. The rate of relapses per year and their duration were not significantly modified in Group 1. The maximum dose of APM required for their relapse control was larger than before treatment although not significantly. All these patients required continuous APM. In Group 2 (IFN treated): 6 patients had improved BPRS scores (N = 6, K = 6, p < 0.01) and 5 improved their RSSE scores (n = 5, K = 5, p < 0.05). In 5 patients there were no relapses on their frequency decreased. There was significant reduction in the duration of the relapses (37.8 +/- 14.6 to 20.7 +/- 12.5 days; t = 4.83; d.f. = 9; p = 0.0009) after treatment. Only 3 patients in Group 2 needed continuous APM after a relapse and the maximum dose required for control was significantly less (1281 +/- 527 to 687 +/- 552 chlorpromazine-equivalent mg. per day; t = 5.56, p < 0.001). Comparisons between groups showed advantage for the IFN treated group in the BPRS change, proportion of patients needing continuous APM and integral evaluation. These results indicate that alpha IFN may be useful in the treatment of Schizophrenic Disorder, Paranoid Type.