Evidence-based medicine combines clinical expertise and the best available evidence from systematic research to aid decision making in patient care. Levels of evidence can be graded from I to V, with level I, the strongest, coming from large randomized controlled trials (RCTs). When a definitive RCT has not been performed, or is impracticable or inappropriate, lesser grades of evidence are used. There is level I evidence supporting the treatment of Helicobacter pylori infection in patients with duodenal or gastric ulcers. Prospective RCTs have shown that cure of the infection is associated with ultimate cure of the ulcer diathesis. Therefore, this is a "grade A" recommendation for treatment. In nonulcer dyspepsia, numerous RCTs have yielded conflicting results regarding the benefits of treatment. Although there are methodological problems with many reported studies, there is some evidence (level II at best) to support treatment--a grade B recommendation. In early gastric cancer and gastric mucosa-associated lymphoid tissue lymphoma, the best available evidence supporting treatment of H. pylori infection is of low quality, i.e., levels III and V. Although these carry only grade C treatment recommendations, treatment is safe and carries at least some evidence of efficacy. It is therefore indicated based on the current best available evidence. No evidence exists to support treating the infection in patients receiving long-term proton pump inhibitors for gastroesophageal reflux disease or in patients with any of the nongastrointestinal conditions that have been tentatively linked to H. pylori.