Granule size distribution (GSD) is considered as one of the major intermediate product attributes in roller compaction (RC) process. Quantitative understanding of the impacts of process conditions and ribbon properties on GSD is essential in implementing Quality-by-Design in product development. In this paper, a two-parameter population balance model (PBM) was formulated and fitted to experimental data, obtained from simulated RC granules using lab-scale oscillating milL To establish a predictive approach, two mathematical models were devised to estimate PBM parameters as functions of ribbon properties. As a major novelty, Young's modulus was used as the input material mechanical property in model parameter determination, as well as ribbon solid fraction. Predictive capabilities of the proposed mathematical functions together with developed PBMs were examined by simulating the data that were excluded in model development step. While holding only two fitting parameters, the proposed methods demonstrated great success in predicting experimental GSD data sets. (C) 2020 Elsevier B.V. All rights reserved.