It is well known that drought periods affect coffee plant development, leading to plant death and abortion of developing fruits in case of severe drought. In relation to coffee genetic diversity, several works reported the identification of plants of C. canephora conilon susceptible or tolerant to drought which were analyzed at the physiological level and also used to identify candidate genes underlying stress responses. Even narrow, a genetic diversity for drought tolerance also exist in the species C. arabica. In addition to the identification of undiscovered transcripts, the recent development of low-cost, high throughput next-generation (NGS) sequencing technologies now opens the way to perform expression profiling and to identify gene presenting differential expression patterns by comparing the frequency of reads obtained after sequencing. In order to initiate such kind of approach in coffee, RNAseq approach was performed using (1) roots of C. canephora conilon susceptible (clone 22) or tolerant (clones 14, 73 and 120) to drought grown under greenhouse conditions with (I) or without (NI) irrigation and (2) meristematic tissues from Iapar59 (I59, drought tolerant) and Rubi (R, drought susceptible) cultivars of C. arabica grown under field-grown with (I) or without (NI) irrigation. These data were compared with those of Coffea transcriptome, including the EST sequences from both C. arabica and C. canephora. Electronic northerns produced by these comparisons identified differentially expressed genes between drought-tolerant and -susceptible clones and cultivars. By qPCR experiments, more than 80 candidate genes, that could play a crucial role in the genetic determinism of drought tolerance in coffee plants, were selected. Based on these results, it can be concluded that the abscisic (ABA) signaling pathway (including ABA synthesis and perception) is one of the major molecular determinants that might explain the better efficiency in controlling stomata closure and