We investigate theoretically the stochastic dynamics of Resonance Electronic Energy Transfer (RET), in a bi-dimensional overexcited ensemble of donor and acceptor molecules. We find that, after initial optical excitation of all the donors, the reaction kinetics is well-described by a nonlinear mean-field theory. The latter provides a solid way to define and compute an effective rate of RET, even for disordered samples. We predict that this effective rate scales as R α with R the average distance between individual excited donors and their nearest-neighbor acceptor molecules, and α ∈ [−6, −2] an exponent depending on the spatial distribution of molecular pairs in the sample. Using a kinetic Monte-Carlo approach, we show departures from this macroscopic meanfield description arising from fluctuations and spatial correlations between several molecules involved in the RET process. We expect this prediction to be relevant for both molecular science and biology, where the control and optimization of the RET dynamics is a key issue.