Background: Medicine is increasingly globalised, centralised and concentrated. Central to this are the multinational companies involved in pharmaceutical production, healthcare provision and insurance, and evidence production. This paper provides a sociological perspective on what this means for evidence, drawing upon EBM practitioners' own accounts. Objectives: I aim to move beyond existing sociological critiques of EBM by focusing on the unrecognised commonalities between medical and social scientists. It will analyse how some EBM practitioners are developing their own internal critiques of evidence in the era of globalisation. Methods: The paper analyses postings made on an EBM email discussion list and compares them with key EBM texts. My findings were that while official versions of EBM may reject sociological perspectives, some EBM practitioners have developed sociological understandings of the effects of structural power on both research agendas and data produced. Discussion: These internal critiques move beyond a reliance on accepted EBM methods to prevent bias. Instead, they suggest how social structures produce certain kinds of evidence and rules for interpreting them. I explain how this can undermine the assumption that science is value-free without leading to a relativist position that "anything goes". I suggest how bias can be seen in structural as well as methodological terms through a critique of biomedicine's new power brokers. Conclusions: While sociologists and EBM practitioners may seem to embrace quite different philosophies, this is not necessarily the case. Medical and social scientists have more in common than is sometimes assumed, which can provide a basis for developing effective responses to the globalisation of research and evidence.