Objective. To determine the role of JNK signaling in the development of osteoarthritis (OA) induced by joint injury or aging in mice. Methods. In the joint injury model, 12-week-old wild-type control, JNK1 −/− , JNK2 −/− , and JNK1 fl/fl JNK2 −/− aggecan-Cre ERT2 double-knockout mice were subjected to destabilization of the medial meniscus (DMM) (n = 15 mice per group) or sham surgery (n = 9-10 mice per group), and OA was evaluated 8 weeks later. In the aging experiment, wildtype control, JNK1 −/− , and JNK2 −/− mice (n = 15 per group) were evaluated at 18 months of age. Mouse knee joints were evaluated by scoring articular cartilage structure, toluidine blue staining, osteophytes, and synovial hyperplasia, by histomorphometric analysis, and by immunostaining for the senescence marker p16 INK4a. Production of matrix metalloproteinase 13 (MMP-13) in cartilage explants in response to fibronectin fragments was measured by enzymelinked immunosorbent assay. Results. There were no differences after DMM surgery between the wild-type and the JNK-knockout mouse groups in articular cartilage structure, toluidine blue, or osteophyte scores or in MMP-13 production in explants. All 3 knockout mouse groups had increased subchondral bone thickness and area of cartilage necrosis compared to wild-type mice. Aged JNK-knockout mice had significantly worse articular cartilage structure scores compared to the aged wild-type control mice (mean ± SD 52 ± 24 in JNK1 −/− mice and 60 ± 25 in JNK2 −/− mice versus 32 ± 18 in controls; P = 0.02 and P = 0.004, respectively). JNK1 −/− mice also had higher osteophyte scores. Deletion of JNK resulted in increased expression of p16 INK4a in the synovium and cartilage in older mice. Conclusion. JNK1 and JNK2 are not required for the development of OA in the mouse DMM model. Deletion of JNK1 or JNK2 is associated with more severe age-related OA and increased cell senescence, suggesting that JNK may act as a negative regulator of senescence in the joint.