Background: Rose is one of the most important ornamental crops, but drought limits its yield and quality seriously. Although serval abiotic stress related genes have been identified, the molecular mechanism during rose’s responding to drought stress is still limited. In this study, the transcriptomes of leaves of two-year-old cutting seedlings of Rosa chinensis ’Old Blush’ from three continuous droughted stages (30, 60, 90 days after full watering) and rewatering were analyzed using RNA sequencing. Meanwhile, weighted gene co-expression network analysis (WGCNA) was used to construct a co-expression network associated with the physiological traits of drought response to discover the hub transcription factors involved in drought response.Results: More than 45 million high-quality clean reads were generated from samples and used for comparison with the rose reference genome. A total of 46433 differentially expressed genes were identified. Gene ontology (GO) term enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis indicated that drought stress caused significant changes in signal transduction, plant hormones, primary and secondary metabolism, and a certain degree of recovery after rewatering. Gene co-expression analysis identified 18 modules, in which four modules have highly degree of correlation with physiological traits. In addition, twenty-two transcription factors including members of NACs, WRKYs, MYBs, ERFs, ARFs and bHLHs with high connectivity in these four modules were screened. Conclusions: The results of this study provided the first transcriptome sequencing report of the drought stress in R.chinensis, and also provided candidate transcription factors in response to drought stress, providing clues for improving the drought tolerance of the rose through molecular breeding in the future.