Concordance--that is, shared decision-making between doctors and patients--is nowadays accepted as an integral part of good clinical practice. It is of particular importance in the case of treatments with only marginal benefits such as those recommended in guidelines for the management of common, chronic diseases. However, the implementation of guideline-based medicine conflicts with that of concordance. Studies indicate that patients are not adequately informed about their treatment. Clinical guidelines for conditions such as cardiovascular disease are based on large-scale randomized trials and the complex nature of the data limits effective communication especially in an environment characterized by time constraints. But other factors may be more relevant, notably pressures to comply with guidelines and financial rewards for meeting targets: it is simply not in the interests of doctors to disclose accurate information. Studies show that patients are far from impressed by the small benefits derived from large scale trials. Indeed, faced with absolute risk reductions, patients decline treatment promoted by guidelines. To participate in clinical decisions, patients require unbiased information concerning outcomes with and without treatment, and the absolute risk reduction; they should be told that most patients receiving long-term medication obtain no benefit despite being exposed to adverse drug reactions; furthermore, they should be made aware of the questionable validity of large-scale trials and that these studies may be influenced by those with a vested interest. Genuine concordance will inevitably lead to many patients rejecting the recommendations of guidelines and encourage a more critical approach to clinical research and guideline-based medicine.