Background: The effect of knowledge of Helicobacter pylori eradication rates on physician choice of treatment regimen is unknown. As practice variation results in differences in outcome, it is important to determine whether physician behaviour can be altered by such knowledge. Aims: (i) To determine whether dissemination of practice variation and effectiveness data regarding H. pylori changes subsequent prescribing behaviour and (ii) whether this change results in an improvement in the effectiveness of therapy. Methods: Community gastroenterologists in the Portland metropolitan area enrolled patients being treated for H. pylori. The regimen used, diagnostic method. indication and success in eradication was measured. Patient-centred factors were also measured, including symptoms, interest in post-treatment diagnostic testing and willingness to pay. Results: Significantly more physicians participating in both studies used proton pump inhibitor-triple therapy based regimens in this trial (46% vs. 85%. P = 0.01), although the overall difference between the two trials was not significant (62% vs. 83%, P = 0.11). There was no change in overall eradication rates by per protocol analysis between trials (84% vs. 85%, P = 0.78), but a significant decrease in effectiveness by intention-to-treat analysis observed in this study (80% vs. 71%, P = 0.03). Significantly more patients were treated for reasons other than peptic ulcer disease in this study (P = 0.0003). Conclusions: The overall effectiveness of H. pylori therapy in practice remains good. There has been a shift in the choice of treatment regimen and indication for therapy between the time periods of the two studies. Dissemination of treatment data appears to effect prescribing behaviour, but whether it has a beneficial effect on treatment outcome remains unproven.