IN a recent article1, MacCallum and Tanner suggest that the traditional homogeneous reaction rate expression is applicable only to isothermal processes, and that an additional term should be added when the temperature varies over the course of the reaction. This suggestion has several disturbing features. It implies, contrary to logic, that the instantaneous reaction rate depends not only on the present state of the system (the frequency of collisions and relative energies and orientations of all of the system molecules) but also—through dT/dt—on previous and future states. The proposed correction is impossible to calculate, because in a real system temperature cannot be varied holding time constant, and therefore the factor (∂C/∂T)t is devoid of physical meaning; moreover, even for the hypothetical case of a step-function change in temperature, the instantaneous change in concentration must be 0, which eliminates the correction. A particularly far-reaching implication of the suggestion is that essentially all of the nonisothermal reactor modelling and analysis that has taken place over the past 80 years has been fundamentally unsound, by virtue of having been based on equation (1).