For more than a decade, various research groups have tracked the population genetics of the oriental fruit fly, Bactrocera dorsalis (Hendel) in China and neighboring countries using mitochondrial cytochrome c oxidase subunit I (COI) DNA. Although most research has reported high levels of mtDNA variation, to date no efforts have been made to integrate and compare the results from these studies simultaneously. Here, we show that: 1) despite the fact that a large portion of the sampling effort has focused on the Yunnan province beginning in 2005, each subsequent study recovers only a small number of previously sampled haplotypes; 2) new haplotypes of B. dorsalis remain to be found, a projection of new haplotypes versus the number of individuals sampled suggest that sampling the species mtDNA diversity is far from reaching an asymptote; 3) it is unlikely that the observed genetic variation is the result of NUMTs (nuclear mitochondrial DNA), as most differences between haplotypes are silent substitutions; and 4) although all studies employed the 3′ end of COI, the length of COI fragment sequenced differs among studies, making comparisons challenging. Therefore, we offer these results with the caveat that mtDNA diversity might be underestimated in China.