The aim of this study was to develop a model for predicting the properties of biocomposites, based on a PHBV biocomposite with sunflower husks, using statistical optimization methods and regression-based predictive modeling. Examination of the tensile strength value according to the different input production parameters like mold temperature, cooling time, packing pressure and injection rate was subjected to a detailed analysis. Various variants of linear regression models were considered, i.e. effect coding of input variables and the original input parameter values. In developing the predictive models, the significance of the variables used was assessed based on linear correlation analysis with heat map generation, Taguchi SNR response matrix analysis, and ANOVA analysis. Additionally, the interactions between input variables were analyzed. The developed predictive models were also validated using the Leave-One-Out-Cross-Validation method. It was determined that mold temperature and cooling time are the key processing parameters that most strongly influence the tensile strength of PHBV–sunflower husk biocomposites.