<p>This study aims to determine the best model for predicting the age-to-weight variation by sexes of newly bred dog breed using nonlinear functional models such as Von Bertalanffy, Gompertz, Logistic, and Richards. For growth curve prediction, a body weight record of 81 dogs was used from the breeding grounds. To evaluate the best model, coefficient of determination (R²) and residual mean square (RMS) statistics were used. In the Richards, Logistic, Gompertz, and Von Bertalanffy models, the coefficient of determination (R²) for males was 0.9936, 0.9935, 0.9897, and 0.9857, the RMS was 0.7040, 0.9641, 1.2145, and 1.4317, respectively, and for females, the coefficient of determination (R²) was 0.9934, 0.9929, 0.9918, and 0.9889, and the RMS was 0.9010, 0.9030, 0.06, and 1.1317, respectively. From the results, it was concluded that the Richard model used in the study was the most suitable model based on the R² and RMS criteria. In conclusion, the Richard model was used to determine the parameters necessary for individual selection and breed improvement, including the maximum body weight, maximum growth rate, and main growth period of the new breed.</p>