Drought is the most important yield limiting abiotic stress worldwide and could have negative effects on food security. This study was conducted to map yield and yield-contributing trait QTLs under drought stress by association mapping, and to in silico identify drought responsive genes and assay their expression profile. Association analysis was conducted using 181 SSR markers and nine phenotypic traits collected from a structured sorghum population consisting of 96 accessions evaluated under optimum and drought stressed conditions at Assiut University Experimental Farm, Assiut, Egypt. Genomic regions associated with major drought tolerance QTLs were annotated with NCBI and sorghum databases to identify drought responsive genes, which were further transcriptionally analyzed. Significant differences between optimum irrigation and drought-stressed conditions were observed in all measured phenotypic traits. Association analysis revealed 93 marker-trait associations between 71 SSR markers and the nine phenotypic traits with an R2 value ranging from 6.56 to 42.64%. A high degree of linkage disequilibrium (>52%) was identified among markers on different chromosomes, suggesting epistatic interaction. BLASTP annotation of QTL genomic sequences identified 86 gene homologs, from which fifteen genes are drought responsive. All fifteen genes were differentially expressed in drought tolerant and susceptible accessions and their expression levels were highly induced in response to drought stress. The results provide a powerful approach coupling association and in silico analyses to identify gene clusters having a common expression profile with similar function, pointing to an evolutionary mechanism to initiate a quantitative drought stress tolerance mechanism.