The high mortality figures for hepatocellular carcinoma (HCC) are mostly due to high recurrence and metastasis rates. However, the metabolic characteristics of HCC metastasis have not been studied extensively. In this study, we attempted to elucidate the metabolite profile of HCC formation and metastasis through proton nuclear magnetic resonance ((1)H-NMR)-based metabolomics. We first established a hepatocellular carcinoma with lung metastasis (HLM) rat model by exposure to diethylnitrosamine. Fifteen rats were then divided into three groups based on pathologic changes: HCC, HLM, and controls. The metabolite profiles of extracts from tumor tissue were obtained using high-resolution (1)H-NMR. One-way ANOVA was used to compare the metabolite levels among the three groups. Multivariate statistical analysis (specifically, unsupervised principal components analysis and supervised partial least squares discriminant analysis (PLS-DA)) were used for HCC and HLM metabolite profiling and data interpretation. PLS-DA models could discern HCC or HLM rats from normal (control) rats. Tumor tissue from HLM showed changes in glucose, lactate, choline, lipids, and some amino acids such as glycine. The results of this pilot study suggest that alterations in glycolysis and the metabolism of glycine and choline occur during HCC invasion and metastasis.