Conventional Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) are the cornerstone in the initial evaluation of brain tumours. The purpose of this study is to evaluate the contribution of Magnetic Resonance Spectroscopy (MRS) and Perfusion-weighted MRI to distinguish malignant from benign tumours. We included 55 patients diagnosed with single brain tumour by CT and MRI, and final histopathological verification of the tumour type: 25 were low-grade gliomas, 8 anaplastic gliomas, 11 glioblastomas, and 11 solitary metastases. We carried out brain MRS and dynamic perfusion-weighted echoplanar MRI in all cases. Perfusion was assessed in the centre of the lesion and in the area of maximum contrast-enhancement. In MRS, we found significant differences in Choline/Creatine ratios in relation to the tumour type with the highest values in high-grade gliomas and metastases. A Ch/Cr ratio equal or higher than 1.78 predicted malignancy at 80% sensitivity and 73% specificity. We found no significant differences in the relative cerebral blood volume (rCBV) for every type of tumour. The mean rCBV was 1.24 for benign tumours and 1.5 for the malignant ones(1.24 for low-grade gliomas, 1.91 for anaplastic gliomas, 1.03 for glioblastomas, and 1.57 for metastases). We conclude that, individually considered, MRS is superior to Perfusion-weighted MRI in the initial assessment of brain tumours. Perfusion MRI has not demonstrated predictive power to distinguish malignant from benign tumours.