Glucokinase activators are regarded as potent candidates for diabetes treatment, however, in clinical studies on patients with type 2 diabetes, a diminishing efficacy was observed after chronic treatment with them. The mechanism of this reduction has not been elucidated, and whether it is a class effect of glucokinase activators remains inconclusive. Here, we firstly identified a diabetic animal model that shows the diminished efficacy after long-term treatment with MK-0941, a glucokinase activator that exhibited diminished efficacy in a clinical study, and we analyzed the mechanism underlying its diminished efficacy. In addition, we evaluated the long-term efficacy of another glucokinase activator, TMG-123. Goto-Kakizaki rats were treated with MK-0941 and TMG-123 for 24 weeks. The results showed that glycated hemoglobin A1C levels and plasma glucose levels decreased transiently but increased over time with the continuation of treatment in the MK-0941-treated group, while decreased continuously in the TMG-123-treated group. Only in the TMG-123-treated group, higher plasma insulin levels were shown at the later stage of the treatment period. For the mechanism analysis, we conducted a hepatic enzyme assay and liver perfusion study in Goto-Kakizaki rats after chronic treatment with MK-0941 and TMG-123, and revealed that, only in the MK-0941-treated group, the activity of glucose-6-phosphatase was increased, and hepatic glucose utilization was decreased compared to the non-treated group. These data indicate that disruptions in hepatic glucose metabolism are involved in the diminished efficacy of glucokinase activators.