The classical JMAK model of recrystallization kinetics has been widely used to describe the growth of randomly distributed nuclei under constant driving force. These conditions are not satisfied in many systems. The driving force for growth generally varies with time, and the nuclei are usually not uniformly distributed. In this paper, we present a physically motivated alternative model of recrystallization kinetics, which accounts for the variation of driving force due to recovery and the effect of clustering of the nuclei. The model is based on the growth kinetics of initially circular, strain free nuclei in a deformed matrix. The effect of recovery on the recrystallization kinetics is studied in terms of the parameters governing initial stored energy and the decay rate. The model predicts cessation of recrystallization when the stored energy decreases below a certain critical limit. This cessation depends on both the initial stored energy value as well as the recovery time constant. The effect of clustering of nuclei on recrystallization kinetics is analyzed by considering representative volume elements with different nuclei distributions. It is shown that recrystallization kinetics becomes slower with increased clustering. The results of this mean field model are compared to phase field simulations. Graphical abstract: [Figure not available: see fulltext.]. © 2021, The Author(s), under exclusive licence to Springer Science+Business Media, LLC part of Springer Nature.