Edgewise growth rates for Widmanstaetten ferrite and bainite in low alloy steels can be represented with an empirical equation showing proportionality to the square of the supersaturation of the austenite. The proportionality constant has a value in reasonable agreement with the assumption of rate control by carbon diffusion. The growth rates are too low to give a noticeable supersaturation of carbon in the growing ferrite. The experimental B[sub s] for low alloy steels does not seem to be related to the T[sub 0] line, nor does B[sub s] evaluated from the incomplete transformation to bainite for an alloy steel. By assuming rate control by carbon diffusion, the empirical equation can be used to calculate the growth rate under paraequilibrium or no partition, local equilibrium (NPLE) conditions. Experimental growth rates for a similar steel falls in-between. The fact that paraequilibrium does not seem to apply is taken as an indication that the [alpha]/[gamma] interface for Widmanstaetten ferrite and bainite is not of a purely martensitic type.