The mechanical reinforcement of nanocomposites containing nanorods-like fillers such as cellulose nanocrystals (CNCs) is often interpreted by adapting the classical parallel–series model, assuming a simple hyperbolic dependence between the percolation threshold and aspect ratio. However, such assumptions are valid only for nanorods with high aspect ratio and often are misinterpreting the reinforcement obtained at low volume fraction of filler loading. To elucidate this intriguing scenario, we proposed a new approach and validated it by compiling and reinterpreting some of available literature that represent the experimental reinforcement with CNCs. Our approach showed better accuracy, specifically for the cases of CNC nanorods with lower aspect ratio. We conclude that this route permits a more realistic evaluation of the mechanical reinforcement, where a physical parameter accounting the polymer filler association is introduced. © 2017 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2017, 134, 45254.