Members of the genus Thermus family are signifiant producers of secondary metabolites, which are very commonly used in industry. Beside the productivity, it is also important to study the growth curve of each strain. The mathematical models which are commonly used to describe behavior of microbial strains under different physical and chemical conditions can reduce measured data.In this study, the mathematical models which describe only the microbial count were used. For analysis, Verhulst model, von Bertalanffy model and Richards model were chosen.During the cultivation of Thermus aquaticus and Thermus scotoductus strains, optical density (OD) and concentration of microorganisms were measured. The mathematical models were fitted to experimental data. The fitting was made in program MATLAB. The coefficients of models and statistical evaluation of goodness of fit of models were identified from mathematical analyses. Graphs of individual models were plotted with prediction bounds. 95% confidence levels were used to statistical evaluation of goodness of fit of models and prediction bounds.Richards model was evaluted as the most corresponding with experimental data. Similar results were reached using Verhulst model. Von Bertalanffy model was not coresponded with experimental data.