The performance of two immunoturbidimetric modifications for rheumatoid factor (RF) testing, which differ with respect to the means of complement inactivation (heat treatment and inactivation with polyvinyl sulphonate), were compared in serum samples from 87 patients with rheumatoid arthritis (RA) and from 403 healthy subjects. IgM-rheumatoid factor titres were also measured with an enzyme linked immunosorbent assay (ELISA). Both immunoturbidimetric tests gave positive reactions (rheumatoid factor > or = 20 IU/ml) in 74 out of the 87 (85%) RA sera. In cases with high RF concentrations the results after chemical inactivation tended to be slightly higher compared with heat inactivation. In healthy subjects rheumatoid factor was detected in 19/403 (4.7%) sera using heat inactivation and in 22/403 (5.5%) sera with chemical inactivation of complement. Interrun coefficient of variation in the chemical inactivation assay was 4.4%; with the heat inactivation method it was 8.1%. In the ELISA, a marginally better correlation was noted in the results obtained using chemical inactivation. Inactivation of complement by means of polyvinyl sulphonate offers the advantage of easier test performance and better reproducibility, and the results may reflect more accurately true rheumatoid factor concentrations.