Because previous studies showed low levels of IFN-gamma in rheumatoid arthritis (RA) synovial fluid (SF) and synovial tissue (ST) explant supernatants, we assayed RA SF and ST for IL-2 and IL-3-like activity. Using an IL-2 dependent murine CTLL line, 6 of 14 RA SF caused increased thymidine uptake (greater than three times control). The activity was distinct from IL-2 because it was not blocked by antibody to IL-2-R. In addition, IL-2 was not detected (less than 50 pg/ml) in 16 joint samples using an ELISA. Multi-colony-stimulating factor (CSF) activity was measured using two assays that can detect murine IL-3 (mast cell proliferation, and bone marrow CSF). In the mast cell assay, [3H]TdR uptake was 493 +/- 67 cpm for medium, 2,910 +/- 329 cpm in the presence of RA SF (p less than 0.001), 1,246 +/- 156 cpm in the presence of SF from patients with seronegative spondyloarthropathies (p less than 0.001), and 736 +/- 100 cpm in the presence of osteoarthritis SF (p greater than 0.1). In the CSF assay, four of five RA SF and five of five RA ST induced colony formation from bone marrow nonadherent cells. Macrophage colonies were most common, although mixed colonies and granulocytes were occasionally observed. The multi-CSF activity in RA is not due to IL-3 since human rIL-3 was not active in either murine assay, and IL-3 mRNA was not detected in RA synovium. Sephadex column chromatography of RA SF revealed that the mast cell growth factor (approximately 6 x 10(3) mol wt) and the CSF (approximately 40 and 100 x 10(3) mol wt) are distinct. The colony-stimulating aspect of the "IL-3-like" activity in RA SF is likely due to CSF-1 because it is the appropriate mol wt and because the activity was neutralized by specific anti-CSF-1 antibody. Finally, an RIA detected 1.6-25 ng/ml of CSF-1 in RA SF and ST and CSF-1 mRNA was detected in four of five RA synovial tissue samples tested.