The widely accepted negative association between schizophrenia and rheumatoid arthritis (RA) is based on the results of investigations which sought RA in large samples of schizophrenic patients. Using a discharge register, we examined the frequency of schizophrenia in a sample of 5626 RA patients. Appendicitis patients (n = 5330) were used as a comparison group. The cumulative incidence of hospital care with the diagnosis of schizophrenia during 8 years was higher in the RA group (0.64%) than in the appendicitis group (0.47%). Schizophrenia was significantly more common in the RA group than in the appendicitis group among the young. The age-adjusted prevalence of schizophrenia was 0.96% in the RA group and 0.51% in the appendicitis group. Because of this unexpected finding, we examined the incidence of RA and appendicitis among a birth cohort born in 1966. The frequencies of RA and appendicitis among schizophrenic cohort members (n = 76), cohort members with psychiatric diagnosis other than schizophrenia (n = 438), and members without psychiatric diagnosis (n = 10503) were similar. These findings do not support the negative association between schizophrenia and RA. Prolonged institutionalization per se may have been the protective factor against RA in the previous studies. The findings also raise the hypothesis that genes that predispose to schizophrenia provide protection from appendicitis, historically a common cause of mortality.