Microarray-based technologies have enabled comprehensive transcriptome profiling.It is becoming feasible to reconstruct gene transcriptional regulatory networks from microarray data. In this chapter, I oudine a new strategy for reconstructing gene regulatory networks as part of the causal network through which genetic variations influence phenotypes. A central step of phenotype manifestation is gene transcription. The gene expres sion programs encoded in DNA sequences are executed via the network of transcriptional regulation. Thus, the gene regulatory networks can be studied as the causal pathways connect ing genetic loci and phenotypes. Bayesian network modeling combined with genetical genomics methods provides a promising method for inferring the causal network from multiple types of data. The complete causal network should include nodes (variables) representing genetic varia tions (e.g., single nucleotide polymorphisms); environmental factors (experimental conditions); phenotypes; and abundances of RNAs, proteins and other bio-molecules. Currently, the recon struction of the causal network focuses on the interactions between genetic variations, abun dance of mRNAs, and phenotypes because large-scale data are available for these components. Advances in molecular profiling will eventually provide sufficient data for all the bio-molecules and thus enable the reconstruction of the complete causal network underlying complex pheno types. This causal network is important not only for understanding the structure of gene regu latory network, but also because it provides deep insights into the molecular underpinnings of genotype-environment-phenotype relations, which are invaluable for uncovering the genetic basis of diseases and predicting the outcomes of therapeutic interventions.