French Polynesia relies solely on the collection of wild Pinctada margaritifera spat for pearl oyster culture. This was developed to help protect the wild populations from overexploitation, but it is feared that massive spat collection could lead to erosion of genetic diversity both in farmed and wild stocks. Wild and farmed collections of P. margaritifera from four atolls in French Polynesia were genotyped at eight microsatellite loci to determine whether there was a loss of genetic diversity from the wild to adjacent farmed aggregations. The average allelic richness for wild samples was not significantly different from that seen for farmed samples, but there was a significant effect of atoll and locus. Pair-wise genetic differentiation (FST) was not significant between adjacent wild and farmed collections or across atolls. Overall there was no evidence for a loss of genetic variability in farmed oysters. Both farmed and wild individuals analyzed here were adults and could have originated from multiple spawning events in time and space. This could have masked genetic processes linked to recruitment happening at a finer scale. P. margaritifera demonstrates high recruitment variability, but the number of parents contributing to a successful cohort of juveniles recruited on collectors is unknown. Low effective number of breeders and variable recruitment are assumed to be responsible for the genetic patchiness that has been observed at a small spatial scale for this species and this could lead to a loss of genetic diversity in both the farmed and wild stocks. The genetic diversity and family make-up of three groups of 1.5 year old oysters were assessed using 13 microsatellite markers. These individuals were harvested on collectors in three closely located sites of the Takapoto atoll. Higher recruitment density and higher allelic richness was observed in one zone compared to the other two. Significant genetic differentiation was also observed at a small spatial scale. Pair-wise FST estimates between collectors within zone were not significant, but were generally significant across zones. Estimates of effective population size and number of families present for these individuals were larger than expected and suggested that the numbers of parents contributing to the recruits on these collector lines were not limited. Similar results were obtained when assessing monthly cohorts of recruits collected in Takapoto over 5 months with 11 microsatellites. Levels of allelic richness were not significantly different among monthly cohorts, and were comparable to the levels observed in the adult samples above. Small but significant temporal genetic differentiation was observed between the monthly cohorts. Again, there was no evidence for low effective population size or for significant family structuring and it did not appear that a limited number of parents produced these temporal cohorts. Patchy genetic structure was observed, but recruitment on collectors does not seem to be driven by a limited number of successful parents. It does not appear that the current pearl culture practices are negatively impacting the local farmed and wild stocks of P. margaritifera in French Polynesia by reducing their levels of genetic diversity.