A thorough examination of the results of existing models based on the liquid sublayer dryout theory suggested the need to postulate a new mechanism to predict the CHF in subcooled water flow boiling. Considering that we have local boiling with bulk subcooled conditions, there will be a distance from the wall at which the fluid temperature is equal to the saturation value. This distance is called ＇superheated layer＇, and is the only region where a bubble may exist. Because of the accumulation and condensation of the vapour generated from the heated wall, a thin elongated bubble, called a ＇vapour blanket＇, is formed, rising along the near-wall region as vertical distorted vapour cylinders. The CHF is postulated to occur when the vapour blanket replenishes the superheated layer, coming into contact with the heated wall (superheated layer vapour replenishment model). The vapour blanket thickness, assumed to be equal to the bubble diameter at the wall detachment, is independent of the heat flux, depending on physical properties, thermal-hydraulic and geometric parameters. The superheated layer depends on the heat flux, physical properties, thermal-hydraulic and geometric parameters. The heat flux for which the superheated layer is equal to the vapour blanket thickness will be the CHF. The comparison of new model predictions with fusion reactor relevant data (0.1 less than or equal to p less than or equal to 8.4 MPa, 0.3 less than or equal to D less than or equal to 25.4 mm, 0.0025 less than or equal to L less than or equal to 0.61 m, 1 less than or equal to G less than or equal to 90 Mg m(-2) s(-1), 25 less than or equal to Delta T-sub,T-in less than or equal to 255 K) is pretty good, as more than 85% of the 1968 data are predicted within +/-25%, with a standard deviation of +/- 16.6%. Besides, because of its structure, based on the heat balance method, the model is applicable to both peripheral uniformly and non-uniformly heated channels.