A simple methodology for assessing the predictive ability of critical heat flux (CHF) correlations applicable to subcooled flow boiling in a uniformly heated vertical tube is developed. Popular correlations published in handbooks and review articles as well as the most recent correlations are analyzed with the PU-BTPFL CHF database, which contains 29 718 CHF data points. This database is the largest collection of CHF data (vertical upflow of water in a uniformly heated round tube) ever cited in the world literature. The parametric ranges of the CHF database are diameters from 0.3 to 45 mm, length-to-diameter ratios from 2 to 2484, mass velocities from 0.01 × 103 to 138 × 103 kg/m2.s, pressures from 1 to 223 bars, inlet subcoolings from 0 to 347°C, inlet qualities from —2.63 to 0.00, outlet subcoolings from 0 to 305°C, outlet qualities from —2.13 to 1.00, and CHFs from 0.05 × 106 to 276 × 106 W/m2. The database contains 4357 data points having a subcooled outlet condition at CHF. A correlation published elsewhere is the most accurate in both low- and high-mass velocity regions, having been developed with a larger database than most correlations. In general, CHF correlations developed from data covering a limited range of flow conditions cannot be extended to other flow conditions without much uncertainty.