Measurement of serum C-reactive protein (CRP) concentration is useful in monitoring the progress of chronic inflammatory diseases. Rheumatoid factor, by its interaction with the Fc portion of IgG, has the potential to interfere with solid-phase immunoassays for CRP and other serum proteins. To determine the effect of RF on a solid-phase enzyme immunoassay for CRP we compared assays employing whole antisera or F(ab')2 fragments. In 92 sera with RF latex titres ranging from 1/80 to 1/81,920, no correlation was found between RF concentrations and CRP measurements. CRP concentrations measured by use of whole antisera (mean +/- standard error of the mean, 59.7 +/- 5.7 mg/l, n = 92) were lower than those measured with F(ab')2 fragments (62.5 +/- 5 mg/l), indicating that exaggeration of CRP measurements did not occur in RF containing sera under the conditions of the assay. Our results show that in the CRP-ELISA, interference from RF was precluded by the high serum dilutions employed. At lower serum dilutions RF binding was detected. Consequently, in solid-phase enzyme immunoassays at lower serum dilutions the presence of RF may lead to false positive results and exaggerated measurements.