The signals of lactate and lipids partially overlap in single-voxel proton MR spectroscopy (1HMRS), sometimes making them difficult to differentiate in clinical settings. Our aim in this study was to identify lactate and lipids by varying the echo time (TE). We expect that the accurate detection of lactate and lipids will have high diagnostic value in the diagnosis of brain tumors. Following our protocol, we obtained meaningful 1HMRS spectra from 213 patients, including 163 patients with brain tumors, between August 1999 and February 2004. 1HMRS was performed with a TE of 144 ms followed by a TE of 30 ms and/or a TE of 288 ms, if necessary. For the 213 patients, lactate level was "negative" in 47 patients, "positive" in 131 patients, and "strongly positive" in 35 patients. The lipid level was "negative" in 90 patients, "positive" in 56 patients, and "strongly positive" in 67 patients. Based on logistic discriminant analyses of neuro-epithelial tumor WHO grade and lactate and lipid levels, lactate and lipid levels were significant between WHO grades 2 and 3 (P=0.0239) and between grades 3 and 4 (P=0.0347). Lipids are a more significant factor for the discrimination between WHO grades 2 and 3 (P=0.0073) and between grades 3 and 4 (P=0.0048). With our method of varying the TE, it is possible accurately and efficiently to detect lactate and lipids in the brain. We found a significant correlation between lactate and lipid expression and WHO grade of neuro-epithelial tumors.