Fingerprint classification comes to be a relevant guarantee for efficient as well as accurate fingerprint identification, in particular in the case of dealing with one-to-many fingerprint identification. Nevertheless, owing to massive intraclass variability, insignificant inter-class variability, and perturbations, the current fingerprint classification methods still need to enhance the accuracy without increasing the computational cost. In this paper, we introduce a novel method that combines the best extractor of features reported in the literature (Hong08) with multilayer extreme learning machines to maintain the superior classification capability (more than 90%) by simplifying the training time (feasibility for realization in a commercial firmware).