Tea plant (Camellia sinensis) is a typical fluoride (F) hyperaccumulator enriching most F in old leaves. There is association between the risk of fluorosis and excessive consumption of teas prepared using the old leaves. It is meaningful to develop methods for controlling F levels in tea leaves. We generated a comprehensive RNA-seq dataset from tea plants grown at various F levels for different durations by hydroponics, aiming at providing information on mechanism of F metabolism in tea plant. Besides raw reads of the RNA-seq dataset, we present assembled unigenes and aligned unigenes with annotations versus the Gene Ontology (GO) databases, Kyoto Encyclopaedia of Genes and Genomes (KEGG) databases, and Nonredundant (Nr) protein databases with low e-values. 69,488 unigenes were obtained in total, in which 40,894 were given Nr annotations.