Titration of continuous positive airway pressure (CPAP) is performed to determine the CPAP setting to prescribe for an individual patient. A prediction equation has been published that could be used to improve the success rate of CPAP titrations. The goals of this study were: (1) to test the hypothesis that the use of the prediction equation would achieve a higher rate of successful CPAP titrations; (2) to validate the equation as an accurate predictor of the prescribed CPAP setting and determine the factors that influence the accuracy of the prediction equation. A total of 224 patients underwent CPAP titration prior to using the equation, with a starting pressure of 5 cm H(2)O. A total of 192 patients underwent CPAP titration using the equation-predicted CPAP level as the starting pressure (median starting pressure of 8 cm H(2)O [interquartile range 7, 10 cm H(2)O]). The percentage of successful studies, as defined by a 50% decrease in the apnea-hypopnea index (AHI) and a final AHI < or =10 cm H(2)O, increased from 50% to 68% (p<0.001), while the number of patients who were prescribed a CPAP level that had not been tested decreased from 22% to 5% (p<0.001). The equation was not accurate in predicting the prescribed level of CPAP, with only 30.8% of the patients with a prescribed pressure < or =3 cm H(2)O of the predicted pressure. Female gender was the only predictor of a prescribed pressure < or =3 cm H(2)O from the predicted pressure (odds ratio 3.45, 95% confidence intervals 1.67, 7.13, p<0.001). A CPAP prediction equation modestly increases the rate of successful CPAP titrations by increasing the starting pressure of the titration. The equation does not accurately predict the prescribed CPAP level, reaffirming the need for a titration study to determine the optimal prescribed level in a given patient.