INTRODUCTION Noninvasive genetics is becoming an in- creasingly valuable tool in population biol- ogy, but obtaining good DNA from nonin- vasive material is still a challenge. Various methods have been used to extract DNA from faeces (see Eggert et al., 2005 for a re- view) in a large variety of mammals (i.e., brown bear — Taberlet and Bouvet, 1992; dugong — Tikel et al., 1996; seel — Reed et al., 1997; wombat — Banks et al., 2003; bats — Vege and McCracken, 2001; Zinck et al., 2004; Carter et al., 2006). The choice of the extraction method is very important in noninvasive genetics (Eggert et al., 2005) as a non efficient method would furnish costly (many PCR replicates needed) and most probably unreliable results (inaccu- rate consensus genotypes) mainly because of high Allelic DropOut (ADO) and False Allele (FA) rates (for a review on geno- typing errors — see Pompanon et al., 2005). On the contrary, an efficient extraction method should provide reliable consensus genotypes with as few PCR replicates as possible. We recently started a project on the con- servation biology of the lesser horseshoe bat (Rhinolophus hipposideros) in Britta- ny. One aim of this project was to set up reliable noninvasive genotyping methods in this species. We first used the protocol pub- lished by Vege and McCracken (2001) to extract DNA from droppings, but it did not yield DNA of good quality in our species. In Acta Chiropterologica, 9(1): 269–276, 2007 PL ISSN 1508-1109 © Museum and Institute of Zoology PAS Good DNA from bat droppings SEBASTIEN J. PUECHMAILLE 1, 2 , GREGORY MATHY 1 , and ERIC J. PETIT 1, 3 1 Ethologie Evolution Ecologie, UMR CNRS 6552, Université de Rennes I, Station Biologique, 35380 Paimpont, France 2 School of Biological and Environmental Sciences, University College Dublin, Belfield, Dublin 4, Ireland 3 Corresponding author: E-mail: eric.petit@univ-rennes1.fr Amplification of a mitochondrial DNA fragment was used to compare the efficiency of five methods for extracting DNA from bat droppings. The Qiagen DNA Stool Kit, which yielded > 90% mtDNA amplification success, was chosen to extract DNA from 586 samples taken over two years in three French colonies of the lesser horseshoe bat (Rhinolophus hipposideros). Samples, for which mtDNA amplification was successful, were subject to the multiplex amplification of eight microsatellite loci. This resulted in > 95% amplification success over 12,592 PCRs. Allelic dropout (ADO) and false allele (FA) rates were low, and consequently, sample and locus quality indexes (QI) were high. These results demonstrate that large scale noninvasive studies of bat colonies are possible. Key words: Chiroptera, error rate, faecal DNA extraction, microsatellite, mtDNA, noninvasive sampling, Rhinolophus hipposideros