Objective - To describe the characteristics and outcomes of patients starting dialysis treatment for end-stage chronic kidney disease in an emergency setting. Methods - Using data from the Réseau épidémiologie et information en néphrologie (Rein) registry, we included 5,921 patients who began dialysis treatment in 2006 across 15 administrative regions. Emergency care is defined as a first dialysis session performed immediately following an evaluation by a nephrologist due to a life-threatening condition. The initial clinical and laboratory characteristics of the patients, as well as survival rates and the likelihood of being placed on the national kidney transplant waiting list, were compared according to the type of care. The confidence interval threshold was set at 95%. Results - Replacement therapy was initiated on a scheduled basis for 73.6% of patients and on an emergency basis for 26.4%. Comorbidities and disabilities are more common in cases of emergency initiation, while placement on the transplant waiting list, use of erythropoietin (EPO), and creation of a vascular access are less common. It appears that 93.9% of patients starting emergency treatment are on hemodialysis in a major center, whereas 25.2% of patients, in cases of scheduled dialysis, are on hemodialysis outside a center or on home peritoneal dialysis. The 3- and 12-month survival rates are 95.9% [95.3–96.5] and 87.4% [86.3–88.3] for scheduled treatments, and 88.0% [86.3–89.5] and 74.2% [72.0–76.3] for those performed on an emergency basis. While 8.4% of scheduled dialysis sessions are performed on patients already on the waiting list and 30.8% are on the list 12 months after initiation, only 22.1% of patients who started on an emergency basis are on the list at 12 months. Conclusion - Starting dialysis on an emergency basis is associated with poor outcomes. This appears to result from a combination of medical, socioeconomic, and organizational factors. These findings support the development of a medical program that ensures continuity of care. (R.A.)