A genetic map is a necessary first step to guide plant breeding programs and to explore the genome structure by localizing genes involved in important agronomic traits expression. For citrus, some genetic maps have been published but more often they were unsaturated or they employed dominant markers, poorly transferable to other labs. We have initiated a collaborative action between INRA (France), CIRAD (France), IVIA (Spain) and the University of Florida (USA) to generate a reliable and saturated genetic map of the Clementine genome with co-dominant markers localized in coding sequences. For this goal, we have investigated the presence of microsatellite sequences (SSRs) inside EST sequences from different libraries developed for sweet orange and Clementine. Nearly 10% of the ESTs include at least one SSR. By this way we expect to map directly genes whose function is identified by sequence comparison with GenBank or NCBI accessions. A segregating population of 88 individuals was created by INRA at San Giuliano by backcrossing an intergeneric hybrid (Clementine x Poncirus trifoliata) with Clementine. Due to the heterozygosity of the two parents (62% and 84% for Clementine and the intergeneric hybrid respectively), the efficiency of polymorphic EST-SSR markers tends to 45%. We have studied the segregation of more than 200 EST-SSR markers, and complemented with 75 genomic SSR markers previously developed we have constructed a genetic map. (Texte intégral)