Background: in recent years increasing survival rates of neonate with congenital diaphragmatic hernia were reported. We present the results of our collective with regard to outcome and predictors of prognosis (need for ECMO, survival). Patients and methods: The neonates with congenital diaphragmatic hernia treated between December 1997 and June 2001 in the university children's hospital Mannheim were included. All patients with congenital diaphragmatic hernia were treated by a defined algorithm including prenatal pulmonary maturation, surfactant immediately after birth, ,,"gentle ventilation,,", inhaled NO, high frequency ventilation and, if necessary, ECMO. We looked at early predictors of the clinical course, e.g. need for ECMO and survival: Birth weight, oxygenation index, ventilation index, the highest and the lowest arterial oxygen and the lowest carbon dioxide partial pressures during the first 24 hours in the ECMO-centre. Results: Between December 1997 and June 200150 neonates with congenital diaphragmatic hernia were treated (24 inborn, all of them diagnosed prenatally, 26 transfered after birth). Mean 01 was 45,5 cmH(2)O/mmHg (no ECMO: 16,4 cmH(2)O/ mmHg; ECMO: 62,1 cmH(2)O/mmHg) mean VI was 133,8 cmH(2)O " mmHg (no ECMO: 83,9 cmH(2)O x mmHg; ECMO: 156,2 cmH(2)O " mmHg). The survival rate was 84%. According to our algorithm 50% of the patients were treated with ECMO, 78% of the patients treated with ECMO survived. 3 Patients were excluded by the therapy option ECMO, because of contraindications. If the patients need to be treated with ECMO, the predictors of prognosis do not allow to draw conclusions on the clinical course regarding the survival of the patients. Conclusion: Our algorithm allows a good survival rate in patients with congenital diaphragmatic hernia. CDH patients seem to have a benefit from transfer into an ECMO centre as early as possible or prenatally. There is no basis according our experiences to exclude patients with congenital diaphragmatic hernia from ECMO. To evaluate the predictors of prognosis regarding the question, which patients do not require ECMO, which patients have a benefit of ecmo, and which patients will die despite ecmo a nationwide cdh-register could be helpfull.