Obesity and insulin resistance may be associated with elevated plasma concentration of branched-chain amino acids (BCAAs) and impaired BCAA metabolism. However, it is unknown whether the insulin-sensitising effect of long-term exercise can be explained by concomitant change in BCAAs and their metabolism.We included 26 sedentary overweight and normal-weight middle-aged men from the MyoGlu clinical trial, with or without dysglycaemia, for 12 weeks of supervised intensive exercise intervention, including two endurance and two resistance sessions weekly. Insulin sensitivity was measured as the glucose infusion rate (GIR) from a hyperinsulinaemic-euglycaemic clamp. In addition, maximum oxygen uptake, upper and lower body strength and adipose tissue depots (using MRI and spectroscopy) were measured, and subcutaneous white adipose tissue (ScWAT) and skeletal muscle (SkM) biopsies were harvested both before and after the 12 week intervention. In the present study we have measured plasma BCAAs and related metabolites using CG-MS/MS and HPLC-MS/MS, and performed global mRNA-sequencing pathway analysis on ScWAT and SkM.In MyoGlu, men with dysglycaemia displayed lower GIR, more fat mass and higher liver fat content than normoglycaemic men at baseline, and 12 weeks of exercise increased GIR, improved body composition and reduced liver fat content similarly for both groups. In our current study we observed higher plasma concentrations of BCAAs (14.4%, p = 0.01) and related metabolites, such as 3-hydroxyisobutyrate (19.4%, p = 0.034) in dysglycaemic vs normoglycaemic men at baseline. Baseline plasma BCAA levels correlated negatively to the change in GIR (ρ = -0.41, p = 0.037) and [Formula: see text] (ρ = -0.47, p = 0.015) after 12 weeks of exercise and positively to amounts of intraperitoneal fat (ρ = 0.40, p = 0.044) and liver fat (ρ = 0.58, p = 0.01). However, circulating BCAAs and related metabolites did not respond to 12 weeks of exercise, with the exception of isoleucine, which increased in normoglycaemic men (10 μmol/l, p = 0.01). Pathway analyses of mRNA-sequencing data implied reduced BCAA catabolism in both SkM and ScWAT in men with dysglycaemia compared with men with normoglycaemia at baseline. Gene expression levels related to BCAA metabolism correlated positively with GIR and markers of mitochondrial content in both SkM and ScWAT, and negatively with fat mass generally, and particularly with intraperitoneal fat mass. mRNA-sequencing pathway analysis also implied increased BCAA metabolism after 12 weeks of exercise in both groups and in both tissues, including enhanced expression of the gene encoding branched-chain α-ketoacid dehydrogenase (BCKDH) and reduced expression of the BCKDH phosphatase in both groups and tissues. Gene expression of SLC25A44, which encodes a mitochondrial BCAA transporter, was increased in SkM in both groups, and gene expression of BCKDK, which encodes BCKDH kinase, was reduced in ScWAT in dysglycaemic men. Mediation analyses indicated a pronounced effect of enhanced SkM (~53%, p = 0.022), and a moderate effect of enhanced ScWAT (~18%, p = 0.018) BCAA metabolism on improved insulin sensitivity after 12 weeks of exercise, based on mRNA sequencing. In comparison, plasma concentration of BCAAs did not mediate any effect in this regard.Plasma BCAA concentration was largely unresponsive to long-term exercise and unrelated to exercise-induced insulin sensitivity. On the other hand, the insulin-sensitising effect of long-term exercise in men may be explained by enhanced SkM and, to a lesser degree, also by enhanced ScWAT BCAA catabolism. Graphical abstract.