ABSTRACT: Drought is the major abiotic stress threatening maize (Zea mays L.) production globally. Despite recent scientific headway in deciphering maize drought stress responses, the overall picture of key genes, pathways, and co-expression networks regulating maize drought tolerance is still fragmented. Therefore, deciphering the molecular basis of maize drought tolerance remains pertinent. Here, through a comprehensive comparative leaf transcriptome analysis of drought-tolerant hybrid ND476 plants subjected to water-sufficient and water-deficit treatment conditions at flared (V12), tasseling (VT), the prophase of grain filling (R2), and the anaphase of grain filling (R4) crop growth stages, we report growth-stage-specific molecular mechanisms regulating maize drought stress responses. Based on the transcriptome analysis, a total of 3,451 differentially expressed genes (DEGs) were identified from the four experimental comparisons, with 2,403, 650, 397, and 313 DEGs observed at the V12, VT, R1, and R4 stages, respectively. Subsequently, 3,451 DEGs were divided into 12 modules by weighted gene co-expression network analysis (WGCNA), comprising 277 hub genes. Interestingly, the co-expressed genes that clustered into similar modules exhibited diverse expression tendencies and got annotated to different GO terms at different stages. MapMan analysis revealed that DEGs related to stress signal transduction, detoxification, transcription factor regulation, hormone signaling, and secondary metabolites biosynthesis were universal across the four growth stages. However, DEGs associated with photosynthesis and amino acid metabolism; protein degradation; transport; and RNA transcriptional regulation were uniquely enriched at the V12, VT, R2, and R4 stages, respectively. Our results affirmed that maize drought stress adaptation is a growth-stage-specific response process, and aid in clarifying the fundamental growth-stage-specific mechanisms regulating drought stress responses in maize. Moreover, genes and metabolic pathways identified here can serve as valuable genetic resources or selection targets for further functional validation experiments.