Neutralizing antibodies (NAb) against interferon-beta (IFN-beta) develop in about a third of treated multiple sclerosis patients and are believed to reduce therapeutic efficacy of IFN-beta on clinical and MRI measures. The expression of the interferon acute-response protein, myxovirus resistance protein A (MxA) is a sensitive measure of the biological activity of therapeutically applied IFN-beta and of its reduced bioavailability due to NAb. However, MxA may not be operative in the pathogenesis of multiple sclerosis or the therapeutic effect of IFN-beta. Instead, matrix metalloproteinases (MMPs) are increased in brain tissue, CSF and blood circulation of multiple sclerosis patients and function as effector molecules in several steps of multiple sclerosis pathogenesis. One of the molecular mechanisms by which IFN-beta exerts its beneficial effect in multiple sclerosis is reduction of MMP-9 expression and increase of its endogenous tissue inhibitor, TIMP-1. Quantitative PCR measurements of MMP-2 and MMP-9, TIMP-1 and TIMP-2, and MxA were performed in peripheral mononuclear cells from clinically stable multiple sclerosis patients with relapsing remitting disease course after short-term and long-term treatment with IFN-beta. IFN-beta therapy down-regulated the expression of MMP-9 and abolished that of MMP-2 in long-term, but not short-term treated multiple sclerosis, while levels of MxA were increased in both instances. The presence of NAb reversed these effects, i.e. led to reduced MxA and increased MMP-2/MMP-9 expression levels compared with NAb- patients. In contrast, expression of TIMPs in peripheral blood mononuclear cells remained unaffected by IFN-beta therapy and the presence of NAb. While MxA is able to detect the biological action and reduced bioavailability of IFN-beta on the basis of single injections, only MMP-9 shows quantitative correlation with the NAb titre. Together with evidence that an imbalance between MMP and TIMP expression is a crucial pathogenetic feature in multiple sclerosis, these findings support the concept of a significant role of NAb in reducing the therapeutic efficacy of IFN-beta.