Growth curves are critical for the understanding and formulation of breeding plans because they shift in response to selection. Nonlinear functions have been used extensively to represent changes in sizes with age, so that the genetic potentials of animals for growth can be evaluated. This study was undertaken to apply the nonlinear Bayesian mixed effect model to examine the changes in the growth pattern of 3 Turkey genotypes in Nigeria. Growth data of 435 turkey poults over a 20-wk period were evaluated using Logistic, von Bertalanffy and Gompertz growth models with normally or Student’s-t distributed error. The parameter estimates were significantly different in the models by sexes and genotypes. However, the estimates of parameters were similar except for the differences observed among the local turkey genotypes whose values decreased significantly under the Student’s-t distribution. The estimate of the average mature weight (A) in the von Bertalanffy model was the closest to the observed average. The estimate of constant B ranged from 15.62 to 22.17 under the Logistic model but significantly dropped to a range of 0.78 to 4.19 with Gompertz and von Bertalanffy models. The estimate of the rate of maturing (k) varied from 0.06 in von Bertalanffy to 0.26 in the Logistic model. Large estimates of A were generally associated with smaller estimates of k in the von Bertalanffy model. Evaluation of the goodness of fit based on the Deviance Information Criteria showed that the von Bertalanffy model was superior. The differences between functions with respect to average life time absolute growth rate, absolute maturing rate and relative growth rate reflect differences in rate of growth and maturing throughout the growing and maturing phases. In general, these differences are indicative of the differences that exist among the functions with respect to their abilities to fit the actual growth curve of these turkey genotypes and their sexes.