How nice would it be to obtain the size distribution of a nanoparticle dispersion fast and without electron microscope measurements? UV-Vis-NIR spectrophotometry offers a very rapid solution; however, the spectra interpretation can be very challenging and needs to take into account the size distribution of the nanoparticles and agglomeration. This work suggests a Monte Carlo method for rapid fitting UV-Vis-NIR spectra using one or two size distributions starting from a dataset of precomputed spectra based on Mie theory. The proposed algorithm is tested on copper nanoparticles produced with Pulsed Laser Ablation in Liquid and on gold nanoparticles from the literature. The fitted distribution results are comparable with Transmission Electron Microscope results and, in some cases, reflect the presence of agglomeration.