Taro (Colocasia esculenta) is an important crop in Africa, Southeast Asia and the subtropics, as it is used for food and medicine. The aims of this study were to access a large expression datasets of taro and reveal the candidate genes of starch synthesis. As a result, approximately 2.2Gb sequence data of taro transcriptome were obtained by using Illumina HiSeq 2000 platform and assembled into 52,935 contigs with an average length of 588.5bp. Sequence similarity analyses against four public databases (NR, GO, KEGG, KOG) found 17,047 contigs that could be annotated with gene descriptions, conserved protein domains, or gene ontology terms. Among the important metabolic pathways, 26 genes related to starch synthesis were validated by RT-PCR. This transcriptome dataset can serve as an important public information platform for further studies in gene expression, genomics, and functional genomic studies in C. esculenta.