Genome-wide association study (GWAS) has improved our understanding of complex traits, but challenges exist in distinguishing causation versus association caused by linkage disequilibrium. Instead, the transcriptome-wide association study (TWAS) detects direct associations between expression levels and phenotypic variations, providing an opportunity to prioritize candidate genes better. We have established an analytical platform using publicly available databases from Arabidopsis accessions. The feasibility of our TWAS was evidenced by identifying associated genes as formerly known factors related to growth allometry, metabolite production, and flowering time. We also identified and validated six new genes associated with flowering time. To further reveal the genes regulating phosphorus (P) starvation responses, we generated transcriptomes from the roots of 190 Arabidopsis accessions grown in P-replete or P-deplete media. Several P-related traits, such as the shoot phosphate (Pi) content, Pi-uptake activity, and the protein level of Pi transporters, were subjected to association studies. Several well-known Pi starvation regulators were found, which supports the effectiveness of our newly generated databases and analysis platform for TWAS. Notably, several novel candidate genes were uncovered, for which their functional association with P-related traits is currently undertaken. Altogether, our transcriptome-to-phenotype associations will offer a platform to discover potential regulators for any desired traits.