We investigated the possible differences in epidemiology, clinical course, management, and outcome between early and late occurrence of shock using data from the Sepsis Occurrence in Acutely Ill Patients Study, a large European multicenter study, which prospectively collected data from all adult intensive care unit (ICU) patients admitted to a participating center within a 2-week period in 2002. Shock was defined as hemodynamic compromise necessitating the administration of vasopressor agents. Early and late shock were defined as onset of shock within the first 2 days in the ICU or later, respectively. Of 3,147 patients, 1,058 (33.6%) had shock at any time, of whom 462 (43.7%) had septic shock. Patients with late shock had a higher incidence of respiratory (87.4 vs. 69.7%, P < 0.001) and hepatic (15.5 vs. 8.7%, P < 0.05) failure, and more often received dopamine (44.7% vs. 34.5%, P < 0.05) and albumin (31.1% vs. 20.3%, P < 0.001) than patients who developed shock early. Intensive care unit and hospital mortality rates were greater in patients who developed shock late, rather than early (52.4% vs. 36.8% and 55.3% vs. 43%, respectively, P < 0.02). In a multivariable analysis, late shock was associated with an independent risk of higher ICU mortality in shock patients (odds ratio, 2.6; 95% confidence interval, 1.6-4.3, P < 0.001). These observations have important implications in establishing individual prognosis as well as in the design and interpretation of clinical trials.