The understanding of biological ligand–receptor binding processes is relevant for a variety of research topics and assists the rational design of novel drug molecules. Computer simulation can help to advance this understanding, but, due to the high dimensionality of according systems, suffers from the severe computational cost. Based on the framework provided by conformation dynamics and transition state theory, a novel heuristic approach of simulating ligand–receptor binding processes is introduced, which is not dependent on calculating lengthy molecular dynamics trajectories. First, the relevant portion of conformational space is partitioned with meshless methods. Then, each region is sampled separately, using hybrid Monte Carlo. Finally, the dynamical binding process is reconstructed from the static overlaps between the partial densities obtained in the sampling step. The method characterizes the metastable steps of the binding process and can yield the corresponding transition probabilities.