The concept of chronic critical limb ischaemia (CLI) emerged late in the history of peripheral arterial occlusive disease (PAOD). The historical background and changing definitions of CLI over the last decades are important to know in order to understand why epidemiologic data are so difficult to compare between articles and over time. The prevalence of CLI is probably very high and largely underestimated, and significant differences exist between population studies and clinical series. The extremely high costs associated with management of these patients make CLI a real public health issue for the future. In the era of emerging vascular surgery in the 1950s, the initial classification of PAOD by Fontaine, with stages III and IV corresponding to CLI, was based only on clinical symptoms. Later, with increasing access to non-invasive haemodynamic measurements (ankle pressure, toe pressure), the need to prove a causal relationship between PAOD and clinical findings suggestive of CLI became a real concern, and the Rutherford classification published in 1986 included objective haemodynamic criteria. The first consensus document on CLI was published in 1991 and included clinical criteria associated with ankle and toe pressure and transcutaneous oxygen pressure (TcPO(2)) cut-off levels <50 mmHg, <30 mmHg and <10 mmHg respectively). This rigorous definition reflects an arterial insufficiency that is so severe as to cause microcirculatory changes and compromise tissue integrity, with a high rate of major amputation and mortality. The TASC I consensus document published in 2000 used less severe pressure cut-offs (≤ 50-70 mmHg, ≤ 30-50 mmHg and ≤ 30-50 mmHg respectively). The thresholds for toe pressure and especially TcPO(2) (which will be also included in TASC II consensus document) are however just below the lower limit of normality. It is therefore easy to infer that patients qualifying as CLI based on TASC criteria can suffer from far less severe disease than those qualifying as CLI in the initial 1991 consensus document. Furthermore, inclusion criteria of many recent interventional studies have even shifted further from the efforts of definition standardisation with objective criteria, by including patients as CLI based merely on Fontaine classification (stage III and IV) without haemodynamic criteria. The differences in the natural history of patients with CLI, including prognosis of the limb and the patient, are thus difficult to compare between studies in this context. Overall, CLI as defined by clinical and haemodynamic criteria remains a severe condition with poor prognosis, high medical costs and a major impact in terms of public health and patients' loss of functional capacity. The major progresses in best medical therapy of arterial disease and revascularisation procedures will certainly improve the outcome of CLI patients. In the future, an effort to apply a standardised definition with clinical and objective haemodynamic criteria will be needed to better demonstrate and compare the advances in management of these patients.