The algebraic model developed in Part I for prediction of chilling times, where cooling at the product surface is by evaporation as well as convection, was tested. Experimental measurements were made by chilling cylindrical test samples constructed of a food analogue in an air chiller During chilling, the surface of each sample was continually wetted by a wetting agent of known water activity and the relative humidity of the air was controlled. Over the 30 trials, the proposed method predicted measured equilibrium temperature, intercept and slope parameters of semi-log unaccomplished temperature change vs time plots as accurately as could reasonably be expected taking into account data uncertainties (agreement largely within +/-6%, but at worst +/-10%). Chilling times were also predicted sufficiently accurately for many industrial applications. The simple model can be used with confidence for most chilling conditions likely to be encountered in industrial practice and including a(w) values below 1.0. Part 3 considers applications to food materials in which surface water activity may not be constant.