Knowledge of the growth of stocks is essential to study the dynamics of fish populations, and thus to manage fisheries. For several species, it is not possible to carry out a sampling covering the totality of range size. Thus, an under-sampling coupled to a Bayesian approach could be realized for Seriola dumerili (Risso, 1810) to improve the growth parameters. A total of 166 individuals (89 females and 77 males) was monthly collected during one year, from August 2020 to July 2021, along the eastern coast of Tunisia (Mediterranean Sea, on coasts of Monastir, Sousse and Mahdia). The individual ageing data was carried out from the scales under binocular with transmitted light. For each sex, length-weight relationship (LWR; GW: gutted weight ± 1 g; TL: total length ± 0.1 cm) was estimated. For female and male LWR are presented as follow GW = 0.063TL 2.524 ( R 2 = 0.948, p < 0.05) and GW = 0.056TL 2.555 ( R 2 = 0.980, p < 0.05) respectively. For both sexes, the relationship exhibited allometric growth. The use of Markov Chain Monte Carlo chains was adopted to boost the estimation of Seriola dumerili growth parameters for both sexes. The comparison with Gompertz and Logistic growth models revealed the accuracy of the von Bertalanffy growth model. The estimated growth parameters for the population were TL ∞ = 106.06 ± 5.54 cm, K = 0.19 ± 0.01 year –1 . The enhancement of growth estimates for under-sampled will be useful in the stock assessment methods to sustain Seriola dumerili fishery.