In Type 2 diabetes, increased glycogenolysis contributes to the hyperglycaemic state, therefore the inhibition of GP (glycogen phosphorylase), a key glycogenolytic enzyme, is one of the possibilities to lower plasma glucose levels. Following this strategy, a number of GPis (GP inhibitors) have been described. However, certain critical issues are associated with their mode of action, e.g. an impairment of muscle function. The interaction between GP and the liver glycogen targeting subunit (termed G(L)) of PP1 (protein phosphatase 1) has emerged as a new potential anti-diabetic target, as the disruption of this interaction should increase glycogen synthesis, potentially providing an alternative approach to counteract the enhanced glycogenolysis without inhibiting GP activity. We identified an inhibitor of the G(L)-GP interaction (termed G(L)-GPi) and characterized its mechanism of action in comparison with direct GPis. In primary rat hepatocytes, at elevated glucose levels, the G(L)-GPi increased glycogen synthesis similarly to direct GPis. Direct GPis significantly reduced the cellular GP activity, caused a dephosphorylation of the enzyme and decreased the amounts of GP in the glycogen-enriched fraction; the G(L)-GPi did not influence any of these parameters. Both mechanisms increased glycogen accumulation at elevated glucose levels. However, at low glucose levels, only direct GPis led to increased glycogen amounts, whereas the G(L)-GPi allowed the mobilization of glycogen because it did not block the activity of GP. Due to this characteristic, G(L)-GPi in comparison with GPis could offer an advantageous risk/benefit profile circumventing the potential downsides of a complete prevention of glycogen breakdown while retaining glucose-lowering efficacy, suggesting that inhibition of the G(L)-GP interaction may provide an attractive novel approach for rebalancing the disturbed glycogen metabolism in diabetic patients.