OBJECTIVE: Describe the clinical and epidemiologic characteristics of patients with systemic lupus erythematosus (SLE) admitted to the intensive care unit (ICU). METHODS: a retrospective study with medical records review of patients with systemic lupus erythematosus (SLE) admitted to the ICU between 2004 and 2015 were included. Qualitative variables were described using absolute and relative frequencies. For quantitative variables mean value and standard deviation (SD) or median value with the interquartile range (IQR) depending on data distribution. To compare groups, it was used the Student t-test or Mann Whitney U test as appropriate and Fisher's exact test. RESULTS: 33 patients were included, with a total of 45 ICU admissions, 29 (87.9%) were females with a median age of 26 years. The median time of diagnosis of SLE was two years, (IQR 1.5-5). The most common SLE manifestation and comorbidity were renal disease and hypertension with 27 (81.8%) and 14 (42.4%) respectively. The main reason for admittance was lupus flare with 25 events (55.5%). Infection was the second cause of admission with 19 events (42.2%). The median stay time in the ICU was four days (IQR 2-7). LODS score was 6 (RIQ 5-8), and APACHE II score was 13 (RIQ 11-17.7). There were 29 infections (64.5%) of which 20 (69%) were hospital-acquired. Four (12.1%) patients died. CONCLUSION: Unlike most of the previously reported series, in this study SLE activity was the most common cause of admission in the ICU. A more aggressive disease and difficulties in the ambulatory setting could explain this behavior. Despite the higher percentage of lupus flares, there was lower mortality.