Current immunoassays for interleukin-1 beta (IL-1 beta) are effective for analyzing fluids derived from cultured cells. However, IL-1 beta determinations in human plasma or serum samples are technically complicated by higher protein and lipid concentrations, physicochemical differences which exist between samples from healthy subjects and those experiencing acute phase responses, and by the fact that IL-1 beta can be produced and degraded in the blood collection tube after the sample is drawn. A simple chloroform extraction process has been developed which eliminates several of the interfering factors from plasma samples and increases the amount of IL-1 beta detected by radioimmunoassay and lymphocyte activation assay. In the radioimmunoassay, rabbit sera was found to influence the accuracy and variability of plasma measurements. Improvements in radioimmunoassay reagents and methods are reported which reduce this influence. Finally, different concentrations of IL-1 beta were measured depending on whether serum or plasma was tested. We propose that plasma samples collected with EDTA and aprotinin provide a better determination of free circulating IL-1 beta in vivo than serum samples, which may contain IL-1 beta secreted from blood leukocytes during the clotting process.