Breast cancer metastases have been crucial for treatment and prognosis of breast cancer, therefore, the early diagnosis of metastases is of great significance to improve the patient's survival. The lung is one of the most common metastatic sites of breast cancer. Although there are many techniques used for lung metastasis diagnosis, they have difficulty in detecting lung metastases at an early and resectable stage when metastatic nodules are small or specially located. In this study, we used 1 H NMR based metabonomics in conjunction with multivariate analysis to determine the metabolic phenotypes of lung tissues at different metastatic stages, and identify the common potential biomarkers for early metastasis in both MMTV-PyMT and 4T1 breast cancer lung metastasis models. Multivariate analysis results showed that the increased levels in lactate, alanine, glutamate and creatine were the potential biomarkers at early metastatic stage. With the severity of metastasis, the emerging changes of up-regulated GPC/PC and myo-inositol, together with down-regulated valine were metabolic profiles at late metastatic stage. These findings give a metabonomic overview of breast cancer metastatic lung tissues characterized by a series of changed biochemical pathways including energy metabolism (glycolysis and creatine metabolism), amino acids and phospholipids metabolism, demonstrating metabonomics as a useful approach for identifying biomarkers for tumor metastasis.