To validate predictors of extubation success and failure in mechanically ventilated infants and children by using bedside measures of respiratory function. Prospective, descriptive study. A university-affiliated children's hospital with a 51-bed critical care area. All infants and children who were mechanically ventilated for > or =24 hrs except neonates < or =37 wks gestation and patients with neuromuscular disease. None. Extubation failure was defined as reintubation within 48 hrs of extubation in the absence of upper airway obstruction. Failure rates were calculated for different ranges (selected a priori of preextubation measures of breathing effort, ventilator support, respiratory mechanics, central inspiratory drive, and integrated indices useful in adults. Effort of spontaneous breathing was assessed by the respiratory rate standardized to age, the presence of retractions and paradoxic breathing, inspiratory pressure, maximal negative inspiratory pressure, ratio of inspiratory pressure to maximal negative inspiratory pressure, and tidal volume indexed to body weight of a spontaneous breath. Ventilator support was measured by F(IO2), mean airway pressure, oxygenation index, and the fraction of total minute ventilation provided by the ventilator. Respiratory mechanics was assessed by peak ventilatory inspiratory pressure and dynamic compliance. Central inspiratory drive was assessed by mean inspiratory flow. Frequency to tidal volume ratio and the CROP (compliance, rate, oxygenation, and pressure) indexed to body weight, the integrated indices useful in predicting extubation failure in adults, were also calculated. A regression test for a linear trend in proportions was performed with preselected ranges and the corresponding failure rates. The failure rates from this study (validation group) were compared to those published previously (prediction group) by the chi-square test for proportions. The distribution of categorical variables between groups was analyzed by using the chi-square test or the Fisher's exact test, and p < .05 was considered significant. The study involved 312 patients. There were no differences in any of the clinical characteristics between the prediction and validation groups. The reasons for reintubation were similar in both groups. Preextubation data were also similar between the two groups. There were no differences between the prediction and the validation groups in failure rates with different ranges. There were no differences in the failure rates for any of the cutoff values for peak ventilatory inspiratory pressure, mean airway pressure, F(IO2), oxygenation index, dynamic compliance, tidal volume indexed to body weight of a spontaneous breath, fraction of total minute ventilation provided by the ventilator, and mean inspiratory flow. Bedside measures of respiratory function can predict extubation success and failure in infants and children. Both a low risk and a high risk of failure can be determined by using these measures. Integrated indices useful in adults do not reliably predict extubation success or failure in infants and children. Our study validates our previously published study.