A best evidence topic in cardiac surgery was written according to a structured protocol. The question addressed was whether there is a relationship between hospital or surgeon volume (SV) and postoperative outcome in adult aortic or mitral valve surgery. One hundred and sixty papers were found using the specified search strategy, of which seven papers represented the best evidence to answer this question. The author, journal, date and country of publication, patient group studied, study type, relevant outcomes, methodology scores, study weaknesses and results are tabulated. Outcomes assessed by these studies were variable; four papers used mortality, one paper used morbidity, one paper used care processes and one paper examined all the above-mentioned outcomes. Six papers investigated the effect of hospital volume (HV) on outcome whilst only one paper assessed the effect of both HV and SV on outcome. The type of valve operated on was also mixed; two papers studied aortic valve only, one paper studied mitral valve only and four papers studied both valves. The methodological quality and validity of each study was assessed by a predefined scoring system. The median total quality score was modest and not strong enough to support the conclusions reported by these studies. In addition, volume-outcome relationship can be affected by several factors related to patient, surgeon and hospital. These factors have not been considered in depth by the mentioned papers. However, there may be a positive relationship between hospital procedural volume and mortality which is more likely to be mediated by SV, and there is also a potential relationship with the rate of mitral valve repair and the use of bio-prosthetic valves in elderly patients. We conclude that regionalisation of adult aortic or mitral valve surgery based on such a limited number of modest quality studies would be an indefensible policy. The implementation of such a scheme can have many clinical, practical, economical and political consequences which have not been examined prospectively until today. Furthermore, the relationship between volume and other outcomes rather than mortality needs further assessment.