The discovery a decade ago of the hepatitis C virus (HCV) led to control of post-transfusion hepatitis in many countries, but raised a number of further questions about the transmission and epidemiological distribution of the newly discovered virus. In some countries, the diagnosis of HCV infection has been designated as a notifiable condition, with doctors (or laboratories, or both) required to send case reports to public health authorities. Such case reports of HCV diagnosis are inevitably biased towards people with symptomatic presentations or a known history of exposure, or those who undergo screening for other purposes, such as blood transfusion. Another approach has been the use of serological prevalence surveys either making use of blood that had been taken or tested for other purposes or recruiting specific population groups for the explicit purpose of measuring HCV prevalence. A limited number of surveys have either attempted to obtain true random samples of geographically defined populations, or recruited subjects, such as pregnant women, who may be considered to be broadly representative of the wider population. Injecting drug use is clearly the main mode of HCV transmission in countries that have established comprehensive blood screening, and it remains a major public health challenge. Although mother-to-child transmission is known to occur with a frequency of about 7% there is no proven means of prevention and very little is known about the long-term consequences of maternally transmitted infection. The conclusion from a number of somewhat conflicting studies is probably that HCV infection should not be viewed as a sexually transmissible infection from the perspective of population health strategies. However, people with HCV infection and their sexual partners should minimize the extent of blood contact that may occur through sexual activity.