Type 2 diabetes (T2D), one of the most common diseases in the western world, is characterized by insulin resistance and impaired β-cell function but currently it is difficult to determine the precise pathophysiology in individual T2D patients. Non-targeted metabolomics technologies have the potential for providing novel biomarkers of disease and drug efficacy, and are increasingly being incorporated into biomarker exploration studies. Contextualization of metabolomics results is enhanced by integration of study data from other platforms, such as transcriptomics, thus linking known metabolites and genes to relevant biochemical pathways. In the current study, urinary NMR-based metabolomic and liver, adipose, and muscle transcriptomic results from the db/db diabetic mouse model are described. To assist with cross-platform integration, integrative pathway analysis was used. Sixty-six metabolites were identified in urine that discriminate between the diabetic db/db and control db/+ mice. The combined analysis of metabolite and gene expression changes revealed 24 distinct pathways that were altered in the diabetic model. Several of these pathways are related to expected diabetes-related changes including changes in lipid metabolism, gluconeogenesis, mitochondrial dysfunction and oxidative stress, as well as protein and amino acid metabolism. Novel findings were also observed, particularly related to the metabolism of branched chain amino acids (BCAAs), nicotinamide metabolites, and pantothenic acid. In particular, the observed decrease in urinary BCAA catabolites provides direct corroboration of previous reports that have inferred that elevated BCAAs in diabetic patients are caused, in part, by reduced catabolism. In summary, the integration of metabolomics and transcriptomics data via integrative pathway mapping has facilitated the identification and contextualization of biomarkers that, presuming further analytical and biological validation, may be useful in future T2D clinical studies by identifying patient populations that share common disease pathophysiology and therefore may identify those patients that may respond better to a particular class of anti-diabetic drugs.