ABSTRACT: BACKGROUND:The aim of this study was to assess the anti-inflammatory effects of local cryotherapy in human non-septic knee arthritis. METHODS:In the phase I of the study, patients were randomized to receive either ice (30?min; N?=?16) or cold CO2 (2?min; N?=?16) applied twice during 1 day at an 8-h interval on the arthritic knee. In phase II, 16 other ice-treated arthritic knees according to the same protocol were compared to the contralateral non-treated arthritic knees (N?=?16). The synovial fluid was analyzed just before the first cold application, then 24?h later. IL-6, IL-1?, TNF-?, IL-17A, VEGF, NF-kB-p65 protein, and PG-E2 levels were measured in the synovial fluid and compared before/after the two cold applications. RESULTS:Forty-seven patients were included (17 gouts, 11 calcium pyrophosphate deposition diseases, 13 rheumatoid arthritides, 6 spondyloarthritides). Local ice cryotherapy significantly reduced the IL-6, IL-1?, VEGF, NF-kB-p65, and PG-E2 synovial levels, especially in the microcrystal-induced arthritis subgroup, while only phosphorylated NF-kB-p65 significantly decreased in rheumatoid arthritis and spondyloarthritis patients. Cold CO2 only reduced the synovial VEGF levels. In the phase II of the study, the synovial PG-E2 was significantly reduced in ice-treated knees, while it significantly increased in the corresponding contralateral non-treated arthritic knees, with a significant inter-class effect size (mean difference -?1329 [-?2232; -?426] pg/mL; N?=?12). CONCLUSIONS:These results suggest that local ice cryotherapy reduces IL-6, IL-1?, and VEGF synovial protein levels, mainly in microcrystal-induced arthritis, and potentially through NF-kB and PG-E2-dependent mechanisms. TRIAL REGISTRATION:Clinicaltrials.gov, NCT03850392-registered February 20, 2019-retrospectively registered.