Abstract Between 1994 and 1999, several sampling designs were used to measure the growth of the black-lip pearl oyster, Pinctada margaritifera var. cumingi at different sites in French Polynesia. Using a common statistical method, growth data were analysed for nine sites and showed significant geographic variability. Parameters of the Von Bertalanffy model, fitted on the nine data sets, ranged from 147 to 186.5 mm for the H ∞ parameter and from 0.42 to 0.58 year –1 for the k parameter. The φ parameter (combination of H ∞ and k) provided a ranking classification of growth in these nine sites: growth was low in closed atoll lagoons, such as Takapoto, whereas the island lagoons and ocean habitat supported a very promising shell growth. Calculated on the basis of these models, two parameters of interest for pearl farming were computed: 1) the time for pearl oysters to reach a size of 100 mm (size at which they are suitable for nucleus implantation), which ranged from 21 to 26 months, and 2) the annual shell growth increment at this size (correlated to the rate of nacreous deposition on the pearl), which varied from 19.7 to 31.8 mm·year –1. The combination of these two sets of results demonstrated that the time necessary to produce a comparable pearl varied significantly according to the site. Several hypotheses to explain the differences observed in the growth of P. margaritifera in these nine sites are proposed. The most reasonable would appear to be: 1) the negative effect of high temperature (> 30°C), and 2) the degree of water renewal and food supply around the bivalve. Although this paper demonstrated significant variability in growth performance in the nine investigated sites, further investigations are needed, especially concerning the flesh growth rate. Indeed, this highly relevant parameter, which reflects both bivalve health and the suitability of farming sites, has received very little attention to date.