BACKGROUND: Joint contractures are a major complication in patients with spinal cord injuries. Positioning, stretching, and physical therapy are advocated to prevent and treat contractures; however, many patients still develop them. Joint motion (exercise) is crucial to correct contractures. Transcutaneous carbon dioxide (CO(2)) therapy was developed recently, and its effect is similar to that of exercise. This therapy may be an alternative or complementary approach to exercise. QUESTION/PURPOSES: Using an established model of spinal cord injury in rats with knee flexion contractures, we sought to clarify whether transcutaneous CO(2) altered (1) contracture, as measured by ROM; (2) muscular and articular factors contributing to the loss of ROM; (3) fibrosis and fibrosis-related gene expression in muscle; and (4) the morphology of and fibrosis-related protein expression in the joint capsule. METHODS: Thirty-six Wistar rats were divided into three equal groups: caged control, those untreated after spinal cord injury, and those treated with CO(2) after spinal cord injury. The rats were treated with CO(2) from either the first day (prevention) or 15th day (treatment) after spinal cord injury for 2 or 4 weeks. The hindlimbs of rats in the treated group were exposed to CO(2) gas for 20 minutes once daily. Knee extension ROM was measured with a goniometer and was measured again after myotomy. We calculated the muscular and articular factors responsible for contractures by subtracting the post-myotomy ROM from that before myotomy. We also quantified histologic muscle fibrosis and evaluated fibrosis-related genes (collagen Type 1, α1 and transforming growth factor beta) in the biceps femoris muscle with real-time polymerase chain reaction. The synovial intima’s length was measured, and the distribution of fibrosis-related proteins (Type I collagen and transforming growth factor beta) in the joint capsule was observed with immunohistochemistry. Knee flexion contractures developed in rats after spinal cord injuries at all timepoints. RESULTS: CO(2) therapy improved limited-extension ROM in the prevention group at 2 weeks (22° ± 2°) and 4 weeks (29° ± 1°) and in the treatment group at 2 weeks (31° ± 1°) compared with untreated rats after spinal cord injuries (35° ± 2°, mean difference, 13°; 39° ± 1°, mean difference, 9°; and 38° ± 1°, mean difference, 7°, respectively) (95% CI, 10.50-14.86, 8.10-10.19, and 4.73-9.01, respectively; all p < 0.001). Muscular factors decreased in treated rats in the prevention group at 2 weeks (8° ± 2°) and 4 weeks (14°± 1°) and in the treatment group at 2 weeks (14 ± 1°) compared with untreated rats (15° ± 1°, 4.85-9.42; 16° ± 1°, 1.24-3.86; and 17° ± 2°, 1.16-5.34, respectively; all p < 0.05). The therapy improved articular factors in the prevention group at 2 weeks (4° ± 1°) and 4 weeks (6° ± 1°) and in the treatment group at 2 weeks (8° ± 1°) compared with untreated rats (10° ± 1°, 4.05-7.05; 12° ± 1°, 5.18-8.02; and 11° ± 2°, 1.73-5.50, respectively; all p < 0.05). CO(2) therapy decreased muscle fibrosis in the prevention group at 2 weeks (p < 0.001). The expression of collagen Type 1, α1 mRNA in the biceps femoris decreased in treated rats in the prevention group at 2 and 4 weeks compared with untreated rat (p = 0.002 and p = 0.008, respectively), although there was little difference in the expression of transforming growth factor beta (p > 0.05). CO(2) therapy did not improve shortening of the synovial intima at all timepoints (all p > 0.05). CO(2) therapy decreased transforming growth factor beta immunolabeling in joint capsules in the rats in the prevention group at 2 weeks. The staining intensity and Type I collagen pattern showed no differences among all groups at all timepoints. CONCLUSION: CO(2) therapy may be useful for preventing and treating contractures after spinal cord injuries. CO(2) therapy particularly appears to be more effective as a prevention and treatment strategy in early-stage contractures before irreversible degeneration occurs, as shown in a rat model. CLINICAL RELEVANCE: Our findings support the idea that CO(2) therapy may be able to improve the loss of ROM after spinal cord injury.