Skeletal muscle insulin resistance is a central node in metabolic dysfunction [1]. While much is known about the mechanisms underpinning insulin action in muscle there is a paucity of pharmacological approaches to reverse muscle insulin resistance. Here we utilise natural genetic variation in insulin sensitivity to identify novel causal genes and potential therapeutics. Diversity Outbred Australia (DOz) are an outbred population of mice that exhibit 50-fold variation in whole-body insulin sensitivity. We performed metabolic phenotyping and quadriceps muscle proteomic analysis on 240 mice, built linear models and performed linkage analysis to identify potentially causal proteins driving insulin resistance. We then took advantage of Connectivity Map [2] to convert our molecular signature of insulin resistance into a list of candidate small molecules that replicate changes in protein expression that we observed in muscle protein levels in DOz mice. To validate these candidates, we performed GLUT4 translocation assays in GLUT4-HA-L6 cells and considered the ability of each compound to 1) induce GLUT4 translocation independently of insulin; 2) potentiate a submaximal dose of insulin; and 3) reverse palmitate-induced insulin resistance. We then constructed a scoring matrix from both in vitro and in silico data to rank each molecule. Of note, two high scoring candidates, disulfiram [3], and ketoconazole [4], have previously shown potential as metabolic therapeutics. In contrast, our top scoring compound thiostrepton is relatively understudied but almost completely restored insulin stimulated GLUT4 translocation in insulin resistant myotubes. Further research is needed to characterise the mode of action of thiostrepton and determine its in vivo and ex vivo capabilities.