A need for prediction of the microstructure heterogeneity and for the evaluation of the uncertainty of the predictions was the motivation for this project. Deterministic models describing recrystallization and phase transformations during hot deformation and controlled cooling of steel products were replaced by stochastic models. These models account for the random character of the nucleation during phase changes. Equations describing probability of recrystallization [1] and nucleation during phase transformations were proposed on the basis of the metallurgical knowledge. Various mechanisms of the nucleation, including grain boundary and autocatalytic nucleation, were considered depending on the type of the transformation. In consequence, instead of the average values, distributions (histograms) of selected microstructural parameters could be calculated. Identification of the models was performed using inverse analysis for the hot compression and dilatometric tests. The objective function in the inverse analysis composed two parts. The first was the mean square root error (MSRE) between measured and calculated average start/end temperatures of transformations and phase fractions in the constant cooling rate (CCT) tests and between measured and calculated times to the start/end of transformations in the isothermal (TTT) tests. The second was the distance between measured and calculated histograms of the austenite grain size during hot deformation and ferrite grain size after cooling. Earth Movers Distance (EMD) was used as a metric. The models were coupled with the finite element program describing hot rolling and controlled cooling of the steel flat products. Simulations of various processing routes were performed and compared. Histograms of the ferrite grain size after cooling were calculated. Moreover, uncertainty of the process parameters was accounted for and uncertainty of the predictions of the final microstructure was evaluated. Following this, conventional equations describing relations between microstructure and mechanical properties were used to asses uncertainty of the latter. References [1] K. Klimczak, P. Oprocha, J. Kusiak, D. Szeliga, P. Morkisz, P. Przybyłowicz, N. Czyżewska, M. Pietrzyk, Inverse problem in stochastic approach to modelling of microstructural parameters in metallic materials during processing, Mathematical Problems in Engineering, 2022, 9690742.