There is no established immunological marker of multiple sclerosis activity, which reflects the poor understanding of the immunopathogenesis of multiple sclerosis. Passive measurement of the levels of soluble inflammatory markers, whose half lives are usually measured in minutes and hours, can only indicate the extent of instantaneous inflammation, which is known to fluctuate in multiple sclerosis. We favour measurement of immune responses in vitro. As healthy individuals have T cell reactivities to myelin proteins that are postulated to be pathogenic in multiple sclerosis, we prefer non-antigen specific mitogen and recall antigen assays as immunological markers. We illustrate their use in the treatment of 27 patients with multiple sclerosis using a pulse of humanised anti-lymphocyte (CD52) antibody that caused prolonged T cell depletion. The mitogen-induced proliferation, and secretion of IFN-gamma, from peripheral blood mononuclear cells in vitro was significantly reduced after treatment, suggesting that immune responses had been modulated. Such observations will only gain credence as an outcome measure if they are shown to correlate with clinical or radiological measures of multiple sclerosis activity. Perhaps more importantly, aspects of the pathogenesis of multiple sclerosis may be revealed by close immunological surveillance of patients undergoing experimental treatments.