Evidence regarding the relationship between outcome and the number (volume) of patients treated at individual hospitals or by individual surgeons is reviewed and the interplay of other factors such as hospital characteristics, population profiles and referral preferences examined. An inverse relationship between mortality rate and hospital volume has repeatedly been found and, while there have been similar findings for surgeon volume, these results have been less consistent. What is certain is that wide variation in outcome does occur. What is less clear is whether the relationship to volume is a causal one or whether it is due to other factors such as those mentioned above. Despite there being a great deal that we do not understand about these relationships, considerable action has been taken as a result of the studies reported here, in the USA in particular. This has taken the form of rationalization of services, publication of hospital mortality rates and the setting of minimum numbers of specific procedures that should be performed each year by individual surgeons. Understanding of this area should be much greater before rationalization is considered in the name of higher quality and before mortality rates according to hospital or surgeon are published.