AtmiRNET is a novel resource geared towards plant scientists for reconstructing regulatory networks of Arabidopsis miRNAs. By means of highlighted miRNA studies in target recognition, functional enrichment of target genes, promoter identification, and detection of cis- and trans- elements, AtmiRNET allows users to explore mechanisms of transcriptional regulation and miRNA functions in Arabidopsis thaliana, which are rarely investigated so far. High-throughput next generation sequencing (NGS) datasets from transcriptional start sites (TSSs)-relevant experiments as well as five core promoter elements were collected to establish the SVM-based prediction model for Arabidopsis miRNA TSSs. Then, high-confidence transcription factors (TFs) participate in transcriptional regulation of Arabidopsis miRNAs are provided based on statistical approach. Furthermore, both experimentally verified and putative miRNA-target interactions (MTIs), whose validity was supported by the correlations between the expression levels of miRNAs and their targets, are elucidated for functional enrichment analysis. The inferred regulatory networks give users an intuitive insight into the pivotal roles of Arabidopsis miRNAs through the crosstalk between miRNA transcriptional regulation (upstream) and miRNA-mediate (downstream) gene circuits.