Resistance of Helicobacter pylori to antibiotics included in current regimens used to eradicate H. pylori is a major reason for failure. The definition of resistance is not simple, and the clinical relevance of in vitro results must be considered. The different methods of testing antibiotics cannot apply in all cases. Resistance to clarithromycin has a low prevalence rate (< 10%) and its mechanism is well defined (point mutation on the 23S rRNA genes, and decreased binding of the antibiotics to the ribosome). Its clinical relevance is not questioned and, because of a clear occurrence of a bimodal strain population, the method for detecting resistance is not crucial. Resistance to nitroimidazoles is much more common, probably in the range of 30% or more in Europe. Neither the mechanism of action of metronidazole resistance nor its mechanism of is well known. The redox potential inside the cell which is important in reducing metronidazole to its active metabolite is probably a key element, but the exact metabolites involved are not yet known. Metronidazole resistance was found to be clinically relevant when standard triple therapy was used. The relevance is questioned for triple therapies including a proton pump inhibitor, clarithromycin and metronidazole. More clinical data are needed in this field and the use of agar dilutions is recommended to assess the susceptibility of H. pylori to metronidazole. The mechanism of resistance to quinolones has been described but these compounds are not currently used for H. pylori infection. No resistance has yet been described for amoxycillin but continuous surveillance is needed in order to detect new cases, as was recently the case for tetracycline resistance.