A high-protein, low-carbohydrate diet has been regarded as a dietary intervention for weight loss in the obese population. We integrated metabolomics profiles and correlation-based network analysis to reveal the difference in metabolism under diets with different protein:carbohydrate ratios. Rats were fed a control diet (moderate-protein moderate-carbohydrate: MPMC; 20 % protein, 56 % carbohydrate) or HPLC diet (high-protein low-carbohydrate: 45 % protein, 30 % carbohydrate) for 6 weeks. The fat content was equal for both diets. HPLC feeding induced weight loss and reduced adipose weight and plasma triglyceride. Compared to the MPMC diet, HPLC significantly increased plasma α-tocopherol, pyruvate, 2-oxoisocaproate, and β-hydroxybutyrate, and reduced linoleate, palmitate, α-glycerophosphate and pyroglutamic acid. The HPLC-associated urinary metabolite profile was signified with an increase in palmitate and stearate and a reduction of citrate, 2-ketoglutarate, malate, and pantothenate. Pathway analysis implicated a significant alteration of the TCA cycle in urine. Biomarker screening demonstrated that individual metabolites, including plasma urea, pyruvate, and urinary citrate, robustly distinguished the HPLC group from the MPMC group. Correlation-based network analysis enabled to demonstrate that the correlation of plasma metabolite was strengthened after the HPLC diet, while the energy-metabolism relatives 2-ketoglutarate and fumarate correlated positively with phenylalanine, methionine, and serine. The correlation network between plasma–urinary metabolites revealed a negative correlation of plasma valine with urinary β-hydroxybutyrate in MPMC rats. In HPLC rats, plasma 2-oxoisocaproate negatively correlated with urinary pyruvate and glycine. This study using metabolomics analysis revealed the systemic metabolism in response to diet treatment and identified the significantly distinct profiles associated with a HPLC diet.