The Guiana Shield is a sub-region of Amazonia, one of the richest areas on earth in terms of species number. It is also one of the most pristine areas and is still largely unexplored. Species number, distribution, boundaries and their evolutionary histories remain at least unclear but most of the time largely unknown. This is the case for most Anurans, a group which is recognized as threatened globally and is disappearing even from pristine tropical forests. Given the pace of forest destruction and the growing concerns about climate change it is urgently necessary to obtain a better estimate of regional biodiversity in Amazonian frogs as well as a better understanding of the origin and distribution of Anuran diversity. Furthermore, given their sensitivity to climatic conditions, amphibians are a good model to investigate the influence of paleoclimatic events on Neotropical diversification which was supposedly the driving force on biotic evolution during Pleistocene in the Guiana Shield. I first test species boundaries in two species Scinax ruber and Rhinella margaritifera. These species are widely distributed, abundant and largely recognized as species complexes. I used an original species delineation method based on the combined use of mitochondrial and nuclear DNA in phylogenetic and phylogeographic analyses. Phylogenetic analyses demonstrated the polyphyly of Scinax ruber and Rhinella margaritifera. These species consist of multiple lineages that may all merit species status. Conflicting signals of mitochondrial and nuclear markers indicated the possibility of ongoing hybridization processes. Phylogeographic analyses added further information in support of the specific status of these lineages. Our results highlight the utility of combining phylogenetic and phylogeographic methods, as well as the use of both mitochondrial and nuclear markers within one study. This approach helped to better understand the evolutionary history of taxonomically complex groups of species. The assessment of the geographic distribution of genetic diversity in tropical amphibian communities can lead to conclusions that differ strongly from prior analyses based on the occurrence of currently recognized species alone. Such studies, therefore, hold the potential to contribute to a more objective assessment of amphibian conservation priorities in tropical areas. Subsequently, I tested if these first results on cryptic species are generalisable, questioning what would potentially be a minimum estimate of the number of cryptic frog species in Amazonia and the Guiana Shield, using mtDNA with multiple complementary approaches. I also combined isolation by distance, phylogenetic analyses, and comparison of molecular distances to evaluate threshold values for the identification of candidate species among these frogs. In most cases, geographically distant populations belong to genetically highly distinct lineages that could be considered as candidate new species. This was not universal among the taxa studied and thus widespread species of Neotropical frogs really do exist, contra to previous assumptions. Moreover, the many instances of paraphyly and the wide overlap between distributions of inter- and intra-specific distances reinforce the hypothesis that many cryptic species remain to be described. In our data set, pairwise genetic distances below 0.02 are strongly correlated with geographical distances. This correlation remains statistically significant until genetic distance is 0.05, with no such relation thereafter. This suggests that for higher genetic distances allopatric and sympatric cryptic species prevail. Based on our analyses, we propose a more inclusive pairwise genetic distance of 0.03 between taxa to target lineages that could correspond to candidate species. Using this approach, we identify 129 candidate species, two-fold greater than the 60 species included in the current study. This leads to estimates of around 170 to 460 frog taxa unrecognized in Amazonia-Guianas. As a consequence the global amphibian decline detected especially in the Neotropics may be worse than realised. The Rhinella margaritifera complex is characterisized by the presence of many cryptic species throughout its wide distribution, ranging from Panama to Bolivia and almost entire Amazonia. French Guiana has long been thought to harbor two species of this group, though molecular data analysed in previous chapters indicated as many as five lineages. I tested whether morphological measurements are correlated or not with genetic data using discriminant analysis and if diagnostic characteristics among the previously determined lineages can be used to describe these new species. This is a novel integrative method which can lead to a facilitation of the description of cryptic species that have been detected by phylogenetic and/or phylogeographic studies. These analyses, combined with published data of other Rhinella species, indicated that two of these lineages represent previously unnamed species. Two of the remaining are allocable to R. margaritifera while the status of the fifth is still unclear because so far it is morphologically indistinguishable from R. castaneotica. Determining if codistributed species responded to climate change in an independent or concerted manner is a basic objective of comparative phylogeography. Species boundaries, histories, ecologies and their geographical ranges are still to be explored in the Guiana Shield. According to the refugia hypothesis this region was supposed to host a forest refugium during climatic oscillations of the Pleistocene but the causes and timing for this have been criticized. We investigated patterns of genetic structure within 18 frog species in the eastern Guiana Shield to explore species boundaries and their evolutionary history. We used mtDNA and nuclear DNA and complementary methods to compare the genetic diversity spatially and temporally. With one exception all the species studied diversified repeatedly within the eastern Guiana Shield during the last 4 million years. Instead of one Pleistocene forest refugium the Guiana Shield has probably hosted multiple refugia during late Pliocene and Pleistocene. Most of these Pleistocene refugia were probably situated on the coast of French Guiana, Amapà, Suriname and Guyana. This diversification likely resulted from forest fragmentation. Many species deserve taxonomic revisions and their ranges to be reconsidered. The local endemism of the Anuran fauna of the Guiana Shield is likely to be much higher and some areas consequently deserve more conservation efforts. Specifically I questioned whether major intraspecific diversification started before the Pleistocene and occurred within the Guiana Shield or ex situ. According to ecological characteristics of the species involved I will test different diversification hypotheses. The consequences on the diversity and the endemism of the Guiana Shield will be explored. My results demonstrate that we have been grossly underestimating local biological diversity in the Guiana Shield but also in Amazonia in general. The order of magnitude for potential species richness means that the eastern Guiana Shield hosts one of the richest frog fauna on earth. In most of the species studied high levels of mtDNA differentiation between populations call for a reassessment of the taxonomic status of what is being recognised as single species. Most species display deep divergence between eastern Guiana Shield populations and Amazonian ones. This emphasizes that the local endemism in the Guiana Shield of these zones is higher than previously recognized and must be prioritised elements taken into account in conservation planning. Nevertheless, a few other species appear widely distributed showing that widespread species do exist. This underlines the fact that some species have efficient dispersal abilities and that the frog fauna of the eastern Guiana Shield is a mixture of old Guianan endemic lineages that diversified in situ mostly during late Pliocene and Pleistocene and more recently exchanged lineages with the rest of Amazonia. Recognizing this strong historical component is necessary and timely for local conservation as these zones are likely to be irremediably modified in the near future.