The role of tumour necrosis factor (TNF)-alpha or cyclo-oxygenase-2 (COX-2) eicosanoids in dystrophinopathies has been evaluated by chronically treating (4-8 weeks) adult dystrophic mdx mice with the anti-TNF-alpha etanercept (0.5 mg/kg) or the COX-2 inhibitor meloxicam (0.2 mg/kg). Throughout the treatment period the mdx mice underwent a protocol of exercise on treadmill in order to worsen the pathology progression; gastrocnemious muscles from exercised mdx mice showed an intense staining for TNF-alpha by immunohistochemistry. In vivo, etanercept, but not meloxicam, contrasted the exercise-induced forelimb force drop. Electrophysiological recordings ex vivo, showed that etanercept counteracted the decrease in chloride channel function (gCl), a functional index of myofibre damage, in both diaphragm and extensor digitorum longus (EDL) muscle, meloxicam being effective only in EDL muscle. None of the drugs ameliorated calcium homeostasis detected by electrophysiology and/or spectrofluorimetry. Etanercept, more than meloxicam, effectively reduced plasma creatine kinase (CK). Etanercept-treated muscles showed a reduction of connective tissue area and of pro-fibrotic cytokine TGF-beta1 vs. untreated ones; however, the histological profile was weakly ameliorated. In order to better evaluate the impact of etanercept treatment on histology, a 4-week treatment was performed on 2-week-old mdx mice, so to match the first spontaneous degeneration cycle. The histology profile of gastrocnemious was significantly improved with a reduction of degenerating area; however, CK levels were only slightly lower. The present results support a key role of TNF-alpha, but not of COX-2 products, in different phases of dystrophic progression. Anti-TNF-alpha drugs may be useful in combined therapies for Duchenne patients.