We evaluated two pressure-recording analytical method (PRAM) software versions (v.1 and v.2) to measure cardiac index (CI) in hemodynamically stable critically ill children and investigate factors that influence PRAM values. The working hypothesis was that PRAM CI measurements would stay within normal limits in hemodynamically stable patients. Ninety-five CI PRAM measurements were analyzed in 47 patients aged 1-168 months. Mean CI was 4.1 ± 1.4 L/min/m(2) (range 2.0-7.0). CI was outside limits defined as normal (3-5 L/min/m(2)) in 53.7% of measurements (47.8% with software v.1 and 69.2% with software v.2, p = 0.062). Moreover, 14.7% of measurements were below 2.5 L/min/m(2), and 13.6% were above 6 L/min/m(2). CI was significantly lower in patients with a clearly visible dicrotic notch than in those without (3.7 vs. 4.6 L/min/m(2), p = 0.004) and in children with a radial arterial catheter (3.5 L/min/m(2)) than in those with a brachial (4.4 L/min/m(2), p = 0.021) or femoral catheter (4.7 L/min/m(2), p = 0.005). By contrast, CI was significantly higher in children under 12 months (4.2 vs. 3.6 L/min/m(2), p = 0.034) and weighing under 10 kg (4.2 vs. 3.6 L/min/m(2), p = 0.026). No significant differences were observed between cardiac surgery patients and the rest of children. A high percentage of CI measurements registered by PRAM were outside normal limits in hemodynamically stable, critically ill children. CI measured by PRAM may be influenced by the age, weight, location of catheter, and presence of a dicrotic notch.