Background Immunoglobulin E (IgE) is highly related to a variety of atopic diseases, and several genome-wide correlation studies (GWASs) have demonstrated the association between genes and IgE. In this study, we conducted the largest genome-wide association study of IgE in a Taiwanese Han population and aimed to elucidate the genetic architecture of IgE. Methods Genome-wide association study was used to discover the association between variants and IgE. Through HLA imputation, we explored the association between HLA alleles and IgE. In order to explore the pleiotropy relationship between IgE and atopic diseases, we performed both global and local genetic correlation analysis. Moreover, we divided our cohort into a training group and a validation group to construct the polygenic risk score (PRS) of IgE and applied it to test the risk of asthma and atopic dermatitis. Results A total of 8 independent variants showed genome-wide significance, and rs147642819 at 6p21.32 was the most significant signal (p= 1.8×10 ^-19). Seven of the loci were replicated successfully after a meta-analysis of the Japanese population. Among all the HLA alleles, HLA-DQB1*03:03 is the most significant allele. The global genetic correlation showed significance between IgE and asthma. The IgE PRS significantly correlated with the total IgE level. Furthermore, the top 10 quantile IgE PRS group had a potentially higher risk for asthma, which was replicated in the Japanese population as well. Conclusions Our study provided a more comprehensive understanding of the impact of genomic variants including complex HLA alleles on serum IgE.