BACKGROUND Most studies of sepsis are from intensive care units (ICUs). We aimed to investigate community-acquired severe sepsis in a broader population, in order to compare patients treated in or outside an ICU . METHODS We performed a 1-year prospective observational study with enrollment of patients from three units; a general ICU, a combined ICU/non-ICU and a medical ward with limited surveillance facilities. Hospital survivors were followed up for 5 years. RESULTS Overall, 220 patients were included, of which 107 received ICU treatment. The majority of abdominal (77%, P = 0.003) and genitourinary (81%, P < 0.001) infections were found in ICU and non-ICU patients, respectively. Time to first antibiotic administration was longer in ICU-patients (median 3.5 vs. 2.0 h in non-ICU patients, P = 0.011). ICU developed more organ dysfunctions than non-ICU patients (P < 0.001), nevertheless supportive therapy with vasoactive drugs and non-invasive ventilation was documented in 22% and 27% of the latter. Median hospital length of stay was 15 vs. 9 days (P = 0.001), and hospital and 5-year mortality rates 35% vs. 16% (P = 0.002) and 57% vs. 58% (P = 0.892) among ICU and non-ICU patients, respectively. Increasing age (HR 1.06 (1.04, 1.07) per year, P < 0.001), not care level during hospitalization (HR 1.19 (0.70, 2.02), P = 0.514), influenced long-term survival. CONCLUSIONS Half of the subjects with community-acquired severe sepsis never received ICU treatment. Still, use of organ supportive therapy outside the ICU was considerable. Hospital mortality was higher, whereas 5-year survival was similar when comparing ICU with non-ICU patients.