Carbon-labelled breath tests were proposed as tools for the evaluation of human liver function 30 years ago, but have never become part of clinical routine. One reason for this is the complex role of the liver in metabolic regulation, making it difficult to provide essential information for the management of patients with liver disease with a single test and to satisfy the hepatology community. As a result, a battery of breath tests have been developed. Depending on the test compound administered, different metabolic pathways (microsomal, cytosolic, mitochondrial) can be examined. Most available data come from microsomal function tests, whilst information about cytosolic and mitochondrial liver function is more limited. However, breath tests have shown promise in some studies, in particular to predict the outcome of patients with chronic liver disease or to monitor hepatic function after treatment. Whilst we await new substrates that can be used to measure liver function in a more valid manner, and large prospective studies to assess the usefulness of available test compounds, the aim of this review is to describe how far we have come in this controversial and unresolved issue.