Brain tumours lack metabolic versatility and are dependent largely on glucose for energy. This contrasts with normal brain tissue that can derive energy from both glucose and ketone bodies. We examined for the first time the potential efficacy of dietary therapies that reduce plasma glucose and elevate ketone bodies in the CT-2A syngeneic malignant mouse astrocytoma. C57BL/6J mice were fed either a standard diet unrestricted (SD-UR), a ketogenic diet unrestricted (KD-UR), the SD restricted to 40% (SD-R), or the KD restricted to 40% of the control standard diet (KD-R). Body weights, tumour weights, plasma glucose, beta-hydroxybutyrate (beta-OHB), and insulin-like growth factor 1 (IGF-1) were measured 13 days after tumour implantation. CT-2A growth was rapid in both the SD-UR and KD-UR groups, but was significantly reduced in both the SD-R and KD-R groups by about 80%. The results indicate that plasma glucose predicts CT-2A growth and that growth is dependent more on the amount than on the origin of dietary calories. Also, restriction of either diet significantly reduced the plasma levels of IGF-1, a biomarker for angiogenesis and tumour progression. Owing to a dependence on plasma glucose, IGF-1 was also predictive of CT-2A growth. Ketone bodies are proposed to reduce stromal inflammatory activities, while providing normal brain cells with a nonglycolytic high-energy substrate. Our results in a mouse astrocytoma suggest that malignant brain tumours are potentially manageable with dietary therapies that reduce glucose and elevate ketone bodies.