The aim of the current study was to investigate the growth pattern of three genetic strains of broiler chickens including Ross 308, Cobb and Arbor Acres by mathematical models. For this purpose, the body weight of 500 broilers for each strain was recorded weekly. Gompertz, Logistic and Richards functions were considered for data fitting. Three functions were compared by adjusted determination coefficient (R2) and root mean square error (RMSE). For all three models, R2 had high values, ranging from 0.987 to 0.999. The difference among the fitted functions by RMSE was significant compared to the R2. The Richards function had more appropriate description for the growth curve of the Cobb strain, because of having the minimum RMSE, 61.57 compared to 85.43 and 66.61, for Gompertz and Logistic functions, respectively. However, the Gompertz function with the maximum R2, and the minimum RMSE, 73.32 and 3237, respectively, was the most suitable function to describe the growth curve of Arbor Acres strain.