Treatment of multiple sclerosis with interferon-beta involves the risk of the development of immunological response to this protein. The response consists in the production of antibodies that bind or neutralise interferon-beta. The binding antibodies are detectable in the majority of the treated patients and do not influence the efficacy of treatment. Simultaneously, large clinical studies provide evidence that the neutralising antibodies reduce and in high titres suppress the effect of interferon, which causes the frequency of exacerbations and the number of new and intensified outbreaks in magnetic resonance imaging to increase. What is more, it may even result in the accelerated progression of disability. The incidence of the neutralising antibodies depends on the preparation as well as the route of administration and concerns from 2 to 42% of the treated patients; it reaches its maximum between the 6^th and the 24^th month of treatment. The neutralising antibodies may be present temporarily (once they disappear, the efficacy of interferon returns to its initial level), yet a high titre predicts their long-term persistence, even after the drug has been discontinued. Various ways of detecting the neutralising antibodies have been devised – from methods that take advantage of a cytopathic effect to those based on the induction of a luciferase reporter gene. Reliable markers of the in vivo activity of interferon-beta that may complement marking of the antibodies are being sought. For many years, the issue of anti-interferon antibodies was accompanied by controversy that resulted from methodological difficulties in studying their clinical utility. At present, in view of the gathered data, European experts unanimously recommend a repeated screening examination of the neutralising antibody titre for patients treated with interferon and considering the results when making therapeutic decisions.