An editorial on June 8 in the Archives of Internal Medicine discusses the problem of publication bias - that is "negative" papers, especially trials, being less likely to make it into the published record. There are a number of reasons for this, from authors not submitting such papers to journals being less likely to publish them. Everyone now agrees that the consequences for the validity of the scientific record are substantial, though the solution is not simple.
A story in the New York Times (free, registration required) discusses the retraction of a paper published in 2008 in the Journal of Bone and Joint Surgery - British Volume (Recombinant human bone morphogenetic protein-2 for grade III open segmental tibial fractures from combat injuries in Iraq. J Bone Joint Surg Br.
There have been plenty of surveys on this, and now a systematic review and meta-analysis has pulled the best ones together (Fanelli D. How Many Scientists Fabricate and Falsify Research? A Systematic Review and Meta-Analysis of Survey Data. PLoS ONE 4(5): e5738. doi:10.1371/journal.pone.0005738 Published: May 29, 2009).