Rose is one of the most fundamental ornamental crops, but its yield and quality are highly limited by drought. The key transcription factors (TFs) and co-expression networks during rose’s response to drought stress and recovery after drought stress are still limited. In this study, the transcriptomes of leaves of 2-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. Weighted gene co-expression network analysis (WGCNA) was used to construct a co-expression network, which was associated with the physiological traits of drought response to discovering the hub TFs involved in drought response. More than 45 million high-quality clean reads were generated from the sample and used for comparison with the rose reference genome. A total of 46433 differentially expressed genes (DEGs) 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 including ABA, auxin, brassinosteroid (BR), cytokinin, ethylene (ET), jasmonic acid (JA) and salicylic acid (SA), primary and secondary metabolism, and a certain degree of recovery after rewatering. Gene co-expression analysis identified 18 modules, in which four modules showed a high degree of correlation with physiological traits. In addition, 42 TFs including members of NACs, WRKYs, MYBs, AP2/ERFs, ARFs, and bHLHs with high connectivity in navajowhite1 and blue modules were screened. This study provides the transcriptome sequencing report of R. chinensis ‘Old Blush’ during drought stress and rewatering process. The study also identifies the response of candidate TFs to drought stress, providing guidelines for improving the drought tolerance of the rose through molecular breeding in the future.