There’s growing demands to comprehensive and precise analysis of protein O-glycosylation in many fields like cancer studies, biopharmaceutical researches and clinical trials. However, the lack of updated and convenient database made it a big obstacle to refer to and store emerging O-glycoprotein data. To solve this problem, an O-glycoprotein repository named OGP was constructed based on a collection of different sourced O-glycoprotein information including existed databases and public datasets (O-Glycbase 6.00 and Simple-cell dataset), and manually extracted data from literatures published since 1998. A total of 9354 O-glycosylation sites and 11633 site-specific O-glycans mapping to 2133 O-glycoproteins were currently recorded in the repository. OGP is the largest O-glycoprotein repository by far. Based on recorded site data, an O-glycosylation site prediction tool was developed. Moreover, the OGP-backed website was already online (http://www.oglyp.org/). There are four functional modules comprised in the website: Statistic Analysis, Database Search, O-Glycosylation Site Prediction and Data Submit. Each modules are specially designed and user-friendly. The first version of OGP and OGP-backed website allow users to obtain various information of O-glycoproteins, such as protein access, sequence, function, site-specific glycan structures, experimental methods, potential glycosylation sites, etc. O-glycosylation information mining can be done in one site efficiently, which will greatly facilitate the researchers on study of O-glycosylation.