A strong negative association between schizophrenia and rheumatoid arthritis (RA), implying low comorbidity, has been found in 12 of 14 previous studies, which we review. To this literature we add two recently acquired data sets encompassing 28,953 schizophrenia patients, only 31 of whom had comorbid RA. Integrating our new data into those of the previous nine studies, which stratified their populations according to psychiatric diagnosis, we obtain a median frequency of RA in schizophrenia populations of 0.09 percent and a mean frequency of 0.66 percent, well below the expected range of 1 percent. These data robustly support prior studies. We also present a meta-analysis evaluating the association between the two diseases by integrating information derived from nine data sets, each furnishing an estimate of the relative risk of RA in schizophrenia patients versus that in other psychiatric patients. We find that the estimated rate of RA among schizophrenia patients is only 29 percent of the corresponding rate in other psychiatric patients. Further, the relative risk of RA in schizophrenia patients versus that in the general population is even less than 29 percent and could be as low as one-third of this value. We present a new hypothesis involving the platelet activating factor system in an effort to account for this negative association and review the suggestions of other investigators toward this end. Finally, we consider the glutamatergic system dysfunction hypothesis of schizophrenia and suggest a possible common pharmacological approach that may ameliorate some of the symptomatology of both schizophrenia and RA.