It is hard to quantify the trace pollutants in the environment without the corresponding reference standards. Structure identifications of unknown organic pollutants are thus of great importance in environmental analysis. As for polybrominated diphenyl ethers (PBDE) with one substituent of methoxyl group, there are 837 congeners, but only 32 standards are commercially available. In this work, an effective method based on gas chromatography coupled with mass spectrometry (GC-MS) was proposed to predict the potential structures of methoxylated polybrominated diphenyl ethers (MeO-PBDEs). The mass fragmentation pattern using SIM mode not only provided the useful information on the substitution position of methoxyl group, the number of Br atoms, but also guaranteed the high sensitivity for trace analysis. Br distribution patterns of the unknown MeO-PBDEs were revealed by a linear regression model with dummy variables which described the retention time relationship between MeO-PBDEs and the corresponding PBDEs on different types of GC columns. This method was successfully used to identify three new MeO-PBDEs metabolites of BDE-28 as 4-MeO-BDE-22, 4'-MeO-BDE-25 and 4-MeO-BDE-31 in the pumpkins. Therefore, the newly developed structure prediction model based on GC-MS behavior is helpful in the evaluation of unknown PBDE metabolites in the environment.