Failure of patient-triggered ventilation in very immature infants may be due to the use of inappropriate triggering systems. Two types of airflow trigger were therefore compared consecutively to an airway pressure (SLE) triggering system. Each comparison was made in 10 infants, < or =28 weeks of gestation. Comparison was made of the delivered volume, trigger performance and blood gases using each system for 1 h. Both comparisons showed that the airflow triggering systems performed better: one (Draeger Babylog 8000) had a higher sensitivity (p < 0.01) and the other (Bird VIP airflow trigger), in which inflation was terminated by sensing a reduction in inspiratory flow, had a lower degree of asynchrony (p < 0.01) and a tendency to deliver higher volumes. These results suggest that triggering systems sensing airflow changes may be superior to those sensing airway pressure changes in very immature infants. The use of a mechanism to synchronize the termination of inflation to the end of the patient's inspiration may offer further advantages.