A cDNA-AFLP approach was used to identify transcript and/or genes specifically expressed in response to drought in tea. Drought was artificially induced and whole genome transcript profiling was done at three different stages-6 days before wilting, 3 days before wilting and at wilting stage of both tolerant and susceptible cultivars, and genetic differences was thus visualized as polymorphisms in the transcriptome. The cDNA-AFLP technique allowed genes and transcripts to be identified in the tolerant genotype (TV-23) whose expression is responsive to drought stress. The cluster analysis revealed two types of clustering-type I separated the tolerant and susceptible cultivar, whereas type II separated the time point of sample and this may be grouped as early and late responsive transcripts. 108 transcript derived fragments were identified as differentially expressed in tolerant genotypes of which 89 sequences could be obtained. Fifty-nine of them showed homology in the public databases. Functional ontology showed genes related to carbohydrate metabolism, response to stress, protein modification process and translation. Cluster I includes five fragments and cluster II includes 25 fragments. Other genes strongly expressed in response to drought in tolerant genotype would help us in identifying and determining the genetic basis of mechanisms involved in conferring drought tolerance in tea.