Equipment for performing laboratory freezing tests on soils has been developed to study the relation between the rates of heave and heat extraction. Laboratory freezing tests are performed on specimens over 500 min tall. In addition, field equipment for automatically recording frost depth and heave has been installed in a pavement. The equipment gives temperature readings for every 50 mm. Temperature gradients at different depths, from laboratory as well as field, are calculated to estimate the rate of liberation of latent heat as a function of time. The calculations indicate that, for the laboratory testing, the heave rate is almost independent of the heat extraction rate for fairly long periods at a constant frost penetration rate. The recorded field data indicate that the heave rate is on average almost constant for relatively long periods of time with a continuous frost penetration, thereby supporting the existence of a maximum heave rate that can not be surpassed persistently. However, heave rate varies greatly if the heat extraction rate often changes between a high and a low level and can for short periods widely exceed the maximum average heave rate possible over long periods. The observed behavior is explained as a consequence of a limited capacity in the redistribution of water to the frost front to supply the heaving.