Tagging data obtained during the research program called MAC are reanalyzed for skipjack tuna (Katsuwonus pelamis), using the von Bertalanffy-Fabens' growth model. Tagging surveys were initiated during 1996-1999 on-board commercial baitboats in Mauritanian and Senegalese waters. The analyses include the Francis' maximum likelihood estimation approaches accounting for growth variability as a function of the expected growth increment. Bearing in mind the generally high negative correlation between estimates of L∞ and K, and instead of fixing L∞ at an arbitrary level, we expressed L∞ as a simple function of K within a meta-analysis framework. Based on previous estimates of growth parameters for skipjack in the world ocean, the fit of this bioenergetic growth equation gave the following relationship: L∞ = 70.607 K – 0.231. An alternative von Bertalanffy-Fabens' growth model was used in which to L∞ was substituted by this function of K. After the comparison of different error structure models, the estimates corresponding to the best fit under the AIC criteria were obtained for K = 0.251 and, after re-introducing this value in the bioenergetic equation, L∞ = 97.258. These results are not in agreement with growth parameter estimates obtained previously in the same part of the Atlantic Ocean but are consistent with the majority of growth estimates obtained by other investigators using tagging, hard part or length-frequency data. The application of the Francis variance model suggests that the variability of the growth increment of skipjack is not constant.