Earlier detection of tumour responses to treatment would facilitate modification of treatment regimens and reduce unnecessary side effects and the costs of ineffective therapy. Anatomical changes following treatment are often slow to manifest and are occasionally misleading. Molecular imaging targeting dysregulated metabolic pathways in tumours can facilitate earlier detection of treatment response. The aim of this study was to directly compare two metabolic imaging techniques that measure different parts of glycolysis, 2-([18F]fluoro)-2-deoxy-D-glucose positron emission tomography ([18F]FDG-PET) and hyperpolarised [1-13C]pyruvate magnetic resonance imaging), for the purpose of detecting early responses to treatment in mouse models of cancer. Two mouse models of lymphoma, subcutaneous EL4 tumours and Eμ-Myc transgenic mice, were treated with etoposide and cyclophosphamide, respectively. In EL4 tumours 24 h after treatment there was a significant reduction in [18F]FDG uptake with no significant change in the hyperpolarised [1-13C]lactate/[1-13]Cpyruvate ratio. While treatment resulted in significant decreases in glucose transporter expression, there were variable amounts of cell death before and after treatment, potentially explaining this discrepancy. In Eμ-Myc mice, reductions of both [18F]FDG uptake and the [1-13C]lactate/[1-13C]pyruvate ratio were observed after treatment. However, the decreases in [18F]FDG uptake in cervical tumours were partially masked by high uptake in surrounding tissues demonstrating the benefit of improved specificity of hyperpolarised [1-13C]pyruvate for detecting the Warburg effect. In two xenograft models of human colorectal and breast adenocarcinoma, a large reduction in hyperpolarised [1-13C]lactate/[1-13C]pyruvate ratio was observed in all tumours 24 hours after treatment with a TRAIL agonist. However, despite treatment inducing widespread apoptosis and long-term remission, [18F]FDG-PET largely failed to detect a response. Measurements of [18F]FDG uptake in disaggregated tumour cells that had been sorted by fluorescence-activated cell sorting demonstrated that inflammatory infiltration or activation was not responsible for failure to detect a response to treatment with [18F]FDG. Furthermore, [1,6 13C2]glucose infusions into tumour bearing mice demonstrated that tumour uptake of [18F]FDG after treatment was not reflective of overall glycolytic flux.